annual report 2005 - asci.tudelft.nl · asci annual report 2005 5 preface with so many things done...

ASCI Annual Report 2005

1

Annual Report 2005

Advanced School for Computing and Imaging

Chapter 1. ASCI and its Research

2

© 2006 Advanced School for Computing and Imaging p/a DUT / EWI P.O. Box 5031 2600 GA DELFT The Netherlands + 31 15 27 88032 Fax: +31 15 27 86632 [email protected] http://www.asci.tudelft.nl ASCI is a Dutch graduate school established in 1993 and accredited by the Royal Netherlands Academy of Arts and Sciences. Research groups of Delft University of Technology, Vrije Universiteit, University of Amsterdam, Leiden University, University Utrecht, University of Twente, University of Groningen, Eindhoven University of Technology and the Erasmus University Rotterdam participate in ASCI.

ASCI Annual Report 2005 3

Contents Preface.................................................................................................................................................................................. 5 Chapter 1 ASCI and its Research ........................................................................................................................................ 7

1.1 About ASCI ........................................................................................................................................................... 7 1.2 Participating Groups.............................................................................................................................................. 7 1.3 ASCI Research Themes ..................................................................................................................................... 10

Chapter 2 Scientific Output along Research Themes ...................................................................................................... 14 2.1 A: High Performance Computing and Computational Science ............................................................................. 14

2.1.1 Contribution of TUD-TNW-tn-qi....................................................................................................................14 2.1.2 Contribution of UL-LIACS ............................................................................................................................15 2.1.3 Contribution of TUD-EWI-me-ce ..................................................................................................................17 2.1.4 Contribution of UvA FdNWI-scs ...................................................................................................................18

2.2 B: Large scale distributed Information Systems and Embedded Systems............................................................ 26 2.2.1 Contribution of UvA-FdNWI-caps.................................................................................................................26 2.2.2 Contribution of UL-LIACS ............................................................................................................................29 2.2.3 Contribution of TUE-EE-dmes .....................................................................................................................31 2.2.4 Contribution of UT- EEMCS-dacs ................................................................................................................35 2.2.5 Contribution of VU – WI...............................................................................................................................39 2.2.6 Contribution of TUD-EWI-ST-pds.................................................................................................................43 2.2.7 Contribution of TUD-me-ce..........................................................................................................................46 2.2.8 Contribution of UvA-FdNWI-csa...................................................................................................................49

2.3 C: Image and Multimedia Sensing, Processing, Interpretation and Visualisation ................................................. 50 2.3.1 Contribution of UG-CS-svcg ........................................................................................................................50 2.3.2 Contribution of TU/e-vis ...............................................................................................................................52 2.3.3 Contribution of TUE-EE-dmes .....................................................................................................................55 2.3.4 Contribution of TUD-mm-ict .........................................................................................................................57 2.3.5 Contribution of TUE-BT-bmia.......................................................................................................................62 2.3.6 Contribution of TUD-TNW-tn-qi....................................................................................................................68 2.3.7 Contribution of UL-LIACS ............................................................................................................................72 2.3.8 Contribution of UvA-FdNWI-ias....................................................................................................................72 2.3.9 Contribution of EUR-RMI-bigr ......................................................................................................................77 2.3.10 Contribution of UU-WI-ics ............................................................................................................................80 2.3.11 Contribution of TUD-LR-frs ..........................................................................................................................84 2.3.12 Contribution of TUD-EWI-mm-cgcc..............................................................................................................85 2.3.13 Contribution of UvA- FdNWI-isis ..................................................................................................................89 2.3.14 Contribution of UL-LUMC-lkeb.....................................................................................................................95 2.3.15 Contribution of UG-CS-IS ..........................................................................................................................100

2.4 D: Multimedia Information Systems................................................................................................................... 103 2.4.1 Contribution of TUD-mm-ict .......................................................................................................................103 2.4.2 Contribution of UVA-FdNWI-isis ................................................................................................................107

Chapter 3 ASCI-wide Events ........................................................................................................................................... 108 3.1 Cooperation within ASCI ................................................................................................................................... 110 3.2 ASCI 2005 Conference ..................................................................................................................................... 114


4


Preface With so many things done and so many different organisations involved it took some time to assemble, but here they are. This is the annual report of the results of ASCI research for 2005. And a busy year it was. As a continuation of a long tradition, in June the eleventh ASCI conference was held at ‘Het Heijderbos’, Heijen. With more than 100 participants to the three day conference including the majority of ASCI's Ph.D student population, the conference was a great success. The focus of the conference was on the use of computer vision and on distributed computing for smart surroundings. Hence, three distinguished keynote speakers were invited to present their insights enhancing the themes: James Crowley from INRIA Grenoble, Richard Kleihorst from Philips Research Eindhoven, and Andrew Herbert from Microsoft Research Cambridge. ASCI takes pride in forming the home base for PhD-students in the area of imaging and computation and variations. One of the measures of success is participation in the yearly ASCI conference, another is in the participation in ASCI's educational program. ASCI organized successfully GNARP2005, the Graduate Network of Applied Research in Parallel systems. The workshop was held in Garderen. Another event was the Medical Imaging Symposium for PhD students held at Erasmus Medical Center Rotterdam. Organised for the first time, the workshop aims at PhD students in the field of Medical Image Computing, Processing and Analysis and related fields. The primary goal of these workshops and symposiums is to bring PhD students together to learn and to exchange experiences in research in an informal and friendly setting with ample opportunities for discussion and feedback. On the formal side of things, ASCI renewed in 2005 its accreditation as a research school with the Royal Netherlands Academy of Arts and Sciences for another 6 year period into 2011. At the same time, the number of associated universities, the number of groups from these universities and the number of Ph.D students of ASCI grew again. The number of PhD students in ASCI passed the 140 mark in 2005. I hope you find the report an interesting read if only by the following key-numbers: 10 PhD degrees were granted, some 550 scientific papers were published in international journals, books and conferences, 39 advanced software packages are described, and some 110 participations in externally funded research programs. This is the first report in the post-Andy era. Andy Tanenbaum who has lead ASCI from the onset for the past 10 years has stepped down to concentrate on his KNAW-professorship. Andy, well done and happy trails in science! For additional information about ASCI, please visit our web page at http://www.asci.tudelft.nl Prof.dr.ir. Arnold W.M. Smeulders - Scientific Director -


6

ASCI Annual Report 2005

7

Chapter 1 ASCI and its Research

1.1 About ASCI ASCI is a national research school on advanced computer and imaging systems. The school was founded in December 1993, and it was approved by the KNAW (Dutch Royal Academy of Sciences) in May 1995. In 2000 the school got its new accreditation for the coming five years. Participants in ASCI are groups from Delft University of Technology, the University of Amsterdam, the Vrije Universiteit, Leiden University and the University of Utrecht; the University of Twente, the University of Groningen, and Eindhoven Technical University have joined ASCI by association agreements. Research within the school can be characterized as applied, experimental and technical computer science, focussed primarily on parallel and distributed systems and processing, as well as the processing of sensor data, image data and other media. With the emphasis on system development, integration of software and hardware, and the processing of sensory information, it directly addresses the needs in high-performance computing and computing intensive applications, with a special emphasis on computational science, and media-oriented applications such as multimedia, medical imaging, computer vision, industrial automation and CAD/CAM. Other important topics are embedded systems and wide-area systems. The school organizes a graduate program and a research program covering all major subjects concerning parallel, distributed, embedded, and real-time systems, performance analysis, image processing, image analysis, image synthesis, sensor interpretation, pattern recognition and computer vision. Every year ASCI organises the Annual ASCI Conference, the scientific meeting place for all participants in ASCI. Another annual activity is the GNARP workshop (GNARP Graduate Network of Applied Research in Parallel systems) which is organized by PhD students and which is a platform for presenting work in progress. There is also a workshop on imaging topics. This workshop is organized every two or three years.

1.2 Participating Groups The following reseach groups participate in ASCI. They are represented together with their abbreviations. For each group the members are listed (situation January 2007). VU-WI-I Vrije Universiteit, Faculty of Sciences, Division of Mathematics and Computer Science, Dept. of

Computer Science http://www.cs.vu.nl/

Prof.dr. A.S. Tanenbaum, Prof.dr. H.E. Bal, Dr. W. de Jonge, Prof.dr. M.R. van Steen,, Dr. C.D. Gamage, Dr.ing. T. Kielmann, Dr. G.E.O. Pierre, Dr. B.J. Overeinder, Dr. R. van Nieuwpoort, Dr. J. Maassen, Dr.ir. H.J. Bos, Dr.B. Crispo, Dr.ir. C. van Reeuwijk, B.C. Popescu M.Sc., Drs M.D. den Burger, Drs. A.M. Dobber, Drs. N. Drost, Drs. T. van der Schaaf, S. Sivasubramanian M.Sc, drs. M. Szymaniak, drs. S. Voulgaris, drs. J.M.S. Wams, drs. M. Wrzesinska; D. Gavidia Simonetti, MSc, M.R. Rieback MS, drs. W.J. de Bruijn, S. Krishnan Nair, Drs. MH.J. Nijhuis, Drs. G. Portokalidis, Drs. J.N. Herder, J.Slowinska

UvA-FdNWI-caps University of Amsterdam, Department of Computer Science,System and Network Engineering

http://www.science.uva.nl/research/sne/ Prof.dr. L.O. Hertzberger, Dr. A. Belloum, Prof.drs. M. Boasson, dr. C.T.A.M. de Laat, C.Erbas

MSc, drs. J.J. van der Ham UvA-FdNWI-ias University of Amsterdam, Faculty of Science, Informatics Institute, Intelligent Autonomous

Systems. http://www.science.uva.nl/research/isla/ Prof.dr. F.C.A. Groen, Prof.dr. D.M. Gavrila, Dr. B.J.A. Kröse, Dr. L. Dorst, Dr. N. Vlassis, Dr.

Z. Zivkovic, drs. W. van der Mark, drs. J.J. Verbeek, drs. J.R. Kok, drs. M.T.J. Spaan, drs. W.P. Zajdel, drs. S. van Gosliga, O. Booij MSc, ir.J.A. van Kleef, M. Hofmann, Th. Noulas MSc., Drs. J.R.J. Nunnink.

UvA- FdNWI-isis University of Amsterdam, Faculty of Sciences, Informatics Institute, Intelligent Sensory

Information Systems Group. http://www.science.uva.nl/research/isis Prof. dr. A.W.M. Smeulders, Dr. R. van den Boomgaard, Dr. M. Worring, Dr.Ing. J.M. Geusebroek, Dr. T. Gevers, Dr. C.J. Veenman, R.F. Aldershoff, Dr. V.T. Pham, Dr. F.J. Seinstra, A. Diplaros M.Sc., Drs. C.G.M. Snoek, ir. G.J. Burghouts, Mw. G.P. Nguyen, S. Mehra M.A., Drs. J.C. van Gemert, T.K. Dang MSc.


8

UvA-FdNWI-scs University of Amsterdam, Faculty of Science, (Informatics Institute/CSP Laboratory), Section

Computational Science. http://www.science.uva.nl/research/scs Prof.dr. P.M.A. Sloot, Dr. G.D. van Albada, Dr. A.G. Hoekstra, Dr. J.A. Kaandorp, Dr. R.G.

Belleman, K. Iskra M.Sc, Drs. M. Scarpa

UvA-FdNWI-csa University of Amsterdam, Faculty of Sicence, Informatics Institute, Computer Systems Architecture Group http://www.science.uva.nl/research/csa/ Prof.dr. C. Jesshope, Dr. P.M.W. Knijnenburg, Dr. A. Pimentel, Drs. T.B. Bernard K. Bousias, Guang Liang MSc, Drs. M. Thompson, Drs Li Zhang

TUD-EWI-isa-pds Delft University of Technology, Faculty of Electrical Engineering, Mathematics & Computer

Science, Parallel and Distributed Systems Group. http://www.pds.ewi.tudelft.nl

Prof.dr. H.J. Sips, dr. D.H.J. Epema, Prof.dr.ir. A.J.C. van Gemund, Prof.dr. C. Witteveen, dr. K. Langendoen, dr. A.G. Baggio, P.J. Garbacki MSc, ir. G.P. Halkes, Haratcherev M.Sc., ir. J. Pietersma, ir. A. Iosup, A.L. Varbanescu MSc, ir. J.D. Mol, O.O. Sonmez

TUD-EWI-me-ce Delft University of Technology, Faculty of Electrical Engineering, Mathematics & Computer

Science, Computer Engineering Group http://ce.et.tudelft.nl

Prof. dr. S. Vassiliadis, Dr. B.H.H. Juurlink, ir. P.J. de Langen TUD-EWI-mm-cgcc Delft University of Technology, Faculty of Electrical Engineering, Mathematics & Computer

Science, Computer Graphics and CAD/CAM Group. http://graphics.tudelft.nl Prof.dr. F.W. Jansen, Dr. W.F. Bronsvoort, ir. F.H. Post, Dr. W. Pasman, M. Koutek, , C.P.

Botha M.Sc., ir.B.Vrolijk, ir. H.A. van der Meiden, L.Zhao, MSc., E.J. Griffith, MSc, ir. G. de Haan

TUD-EWI-mm-ict Delft University of Technology, Faculty of Electrical Engineering, Mathematics & Computer

Science, Department of Mediamatics, Information and Communication Theory Group. http://www.ict.ewi.tudelft.nl

Prof.dr.ir. J. Biemond, Prof.dr.ir. R.L Lagendijk, dr.ir.R.P.W. Duin,Dr. E.A Hendriks, Dr.ir. J.C.A. van der Lubbe,Prof.dr.ir. M.J.T Reinders, Dr. A Hanjalic, dr. L.F.A. Wessels,. dr.ir. D. de Ridder, dr.ir.E.P. van Someren, Dr. E.M. Pekalska, ir. R.J.M. den Hollander, MTD A.I. Deac, ir. P.J.O. Doets, P.Juszczak M.Sc., C. Lai MSc., ir.J.F. Lichtenauer, ir. M. van Staalduinen, ir. J.R. Taal, A. Harol MSc, ir. Th.A. Knijnenburg, S.A. Verzakov MSc, J.Wang, ir. R.J.P. van Berlo, H. Celik MSc, Drs. G.A. ten Holt, B. Kroon, Yunlei Li, Drs W. Meuleman, U. Naci MSc, Ir. M.H. van Vliet, ir. R.P. Westerlaken

TUD-TNW-tn-qi Delft University of Technology, Faculty of Applied Physics, Imaging Science & Technology,

Quantitative Imaging Group http://www.qi.tnw.tudelft.nl/ Prof.dr. I.T. Young, Prof.dr. L.J. van Vliet, Dr. P.P. Jonker, Prof.dr. A.M. Vossepoel, J. Caarls,

drs. F.G.A. Faas, B.E. TuanQuangPham, I.W.O. Serlie,ir.G.O.F. Parikesit, ir. C. van Wijk, B.J. Vermolen

UL-LIACS/LERC Leiden University, Faculty of Mathematics and Natural Sciences, Leiden Institute of Advanced

Computer Science (LIACS) Computer & Software Systems Division- Leiden Embedded Research Center (LERC)

http://www.liacs.nl/research/CS/ Prof.dr. H.A.G. Wijshoff, Prof.dr. E.F. Deprettere, Prof.dr. F.J. Peters, Dr. M.S. Lew, Dr. A.A.

Wolters, Dr. D.P. Huijsmans, , Dr. E.M. Bakker, Dr.ir. T.P. Stefanov, Drs. J.T. Rijsdam, C. Zissulescu-Ianculescu M.Sc., M. Haneda M.Sc., A. Turjan, drs.M.L. Cristea, drs. V. Zivkovic, drs.I.Cimpian, Drs. Hui Li, H. Nikolov MSc, rs.ing. J.H.T. Rohling, Drs. Bin Jiang, J. Lemaitre, Drs. S. Meijer, Drs. A.A.J. Oerlemans

UU-WI-ics Utrecht University, Faculty of Science, Department of Information and Computing Sciences,

Center for Geometry, Imaging and Virtual Environments. http://www.give.nl Prof.dr. M.H. Overmars, Dr. M. van Kreveld, Dr. F. van der Stappen, Dr. R. Veltkamp, Dr. R.W. van Oostrum, Dr. P. Min, Dr. O. Grigore, M.Sc., Drs R. Geraerts, Drs. A. Kamphuis, dipl.ing.

Research Themes


Iris Reinbacher, drs. J.P. van den Berg, drs. E.J. Moet, Dipl.Inform. R.Typke, O.C. Goemans MSc. Ir. R. van Gulik, Drs. M. Bosma, Drs. F.B. ter Haar, Drs.R.H. van Leuken, M. Vahedi

TUD-L&R-frs Delft University of Technology, Faculty of Aerospace Engineering Department of Earth Observation and Space systems (DEOS)

Optical and Laser Remote Sensing Group http://www.lr.tudelft.nl/olrs Dr. B.G.H. Gorte, dr.ir. F. van den Heuvel, T.R. Shah M.Sc., F. Karimi Nejadasl

UG-CS-svcg University of Groningen, Faculty of Mathematics and Physical Sciences, Institute of

Mathematics and Computing Science, Scientific Visualization and Computer Graphics. http://www.cs.rug.nl/informatica/onderzoek/programmas/svcg Prof.dr. J.B.T.M. Roerdink, Dr. G. Vegter, Dr. H. Bekker

UG-CS-IS University of Groningen, Faculty of Mathematics and Natural Sciences, Institute of Mathematics

and Computing Science, Intelligent Systems. http://www.cs.rug.nl/ Prof.dr. N. Petkov, Dr. M.H.F. Wilkinson

TUE-EE-dmes Technische Universiteit Eindhoven, Faculty of Electrical Engineering, Design Methodology for

Electronic Systems http://www.es.ele.tue.nl/es/ Prof.dr.ir. R.H.J.M.Otten, Dr.ir. T. Basten, Prof.dr. H. Corporaal, Dr.ir.M.C.W. Geilen, Prof.dr.ir. G. de Haan, Dr.ir. J.P.M. Voeten, A. Beric M.Sc., S.H. Fatemi M.Sc., S. Stuijk, M. Zhao M.Sc.; Dipl.Ing. O. Florescu, Dipl.Eng. S.V. Gheorghita, A.H. Ghamarian Msc., H.Hu MSc.

TUE-WI-vis Technische Universiteit Eindhoven, Department of Mathematics and Computer Science, Visualization Group,

http://www.win.tue.nl/vis Prof.dr.ir. J.J. van Wijk, dr. R. van Liere, dr.ir. A.J.F. Kok, dr. A.C. Telea, dr.ir. H.M.M. van de Wetering, ir. F.J.J. van Ham, A.J. Pretorius MSc, PD.Eng. S.L. Voinea, ir. D. Reniers

TUE-BMT-bmia Technische Universiteit Eindhoven, Department of Biomedical Engineering

Biomedical Image Analysis http://www.bmi2.bmt.tue.nl/image-analysis/ Prof.dr. B.M. ter Haar Romeny, Prof.dr.F.A. Gerritsen, Dr. L.M.J. Florack, Dr. A. Vilanova Bartroli, dr.ir. J.M. Hofman, H. Bouma, dr.ir. R. Duits, F. Kanters, P. Sereda M.Sc., E. Balmachnova MSc, ir. E.M. Franken, ir. T.H.J.M. Peeters, ir. B. Platel, ir. F. Kanters

UL-LUMC-lkeb Leiden University Medical Center, division of Image Processing, laboratorium voor klinische en Experimentele Beeldverwerking www.lkeb.nl Prof.dr.ir. Johan H.C. Reiber, dr.ir. Faiza Admira Behloul, ir. Johan G. Bosch, dr.ir. B.P.F. Lelieveldt, Isabel Maria Adame MSc, Luca Ferrarini MSc, drs Marijn van Stralen, drs. Avan Suinesiaputra, Drs.A.E.H. Scheenstra

UT-EEMCS-dacs University of Twente, Faculty of Electrical Engineering, Mathematics and Computer Science, Design and Analysis of Communication Systems http://dacs.ewi.utwente.nl prof.dr. B. Haverkort, Dr.ir. G.J. Heijenk,Dr.ir. A. Pras, Dr.ir. P.T. de Boer, Dr.ir. G. Karagiannis, Ir. P. Goering, A. Jehangir MSc., Ir. M. Kamilova, Ir. Fei Liu, A. Remke Dipl.-Inf.

EUR-RMI-bigr Erasmus MC, Departments of Radiology and Medical Informatics, Biomedical Imaging Group Rotterdam http://www.bigr.nl/ Prof.dr. W.J. Niessen, Prof.dr.ir. A. Vossepoel, Dr.Ir. E. Meijering, Dr. J. Veenland, Dr.Ir. H. Vrooman, Dr.ir. Th. van Walsum, Dr. R. Stokking, L. Alic MSc, Drs. F. van der Lijn, E. Rollano MSc, I. Smal MSc


10

1.3 ASCI Research Themes In this report the scientific output of ASCI over the year 2004 has been collected and categorized along the two main research lines of ASCI, Computing and Imaging. These are divided into research themes (A,B,C and D – see matrix) of which the denomination has changed a bit. ASCI research historically comprises two main themes: Computing and Imaging. Within these themes, activities can be divided into methods & algorithms and systems & architecture. Methods & algorithms deals with the development of models and tools for scientific and industrial applications. Systems & architecture deals with the large scale integration in areas like telematics, embedded systems, communication and networks. Both types of activity are targets for fundamental and applied research within ASCI. These themes and target areas make it possible to construct the following matrix, in which four fields can are defined. These fields are indicated as A, B, C and D and are made more specific in the following paragraphs.

Methods & Algorithms

Systems & Architecture

Computing

A High Performance Computing and Computational Science

B Large scale distributed Information Systems and Embedded Systems

Imaging

C Image and Multimedia Sensing, Processing, Interpretation and Visualisation

D Multimedia Information Systems

A: High Performance Computing and Computational Science

A1: High Performance Computing High Performance Computing (HPC) is the collective name of large scale and/or real-time calculations on state-of-the-art computers. These computers are the very tools used in computational science to allow thrusting back the frontiers of knowledge. The Computing branch of ASCI is interested in HPC to develop, improve, and study program models and programming tools for different HPC architectures. Of special interest to ASCI is research on distributed HPC architectures, which will form the basis of future Grid computing. To this purpose ASCI has a research distributed HPC system available: the DAS (Distributed ASCI Supercomputer). The research in distributed HPC comprises applications, algorithm design, languages and compilers, run-time systems, and scheduling techniques.

A2: Computational Science Besides theoretical and experimental research, modeling and simulation has become the powerful third paradigm of scientific inquiry in the natural sciences. Here, models of natural phenomena are analyzed through (in many cases large-scale) simulations executed on state-of-the-art computing system, ranging from desktop workstations to heterogeneous distributed (grid-based) environments or massively parallel computers. Over the years, researchers have realized that modeling and simulation has many generic elements, not connected to a specific application field. Furthermore, it became clear that successful modeling and simulation requires multidisciplinary teams, where computer scientists, numerical mathematicians and researchers from application fields work closely together. These developments led to the evolution of the field of computational science. Here, research focuses on computational methods, models, and tools, using new insight from computer science and (numerical) mathematics, with the goal to facilitate the study of processes, through simulation, that was hitherto not possible. Within ASCI a number of groups are active in the field of Computational Science. Their embedding in ASCI provides necessary input from basic computer science, especially from the computing theme. Computational Science within ASCI is basic research (as opposed to application driven computational science), focusing on generic models, enabling tools and problem solving environments. Generic Models : Capable to capture many natural systems, allowing for formal study (using e.g. concepts from theoretical computer science), and preserving inherent parallelism. Examples include Cellular Automata, particle based models, and

Research Themes


natural solvers. Also, study of fundamental issues related to e.g. synchronous vs. asynchronous execution in generic (parallel) models. Enabling Tools : Examples include Virtual labs, grid based computing environments, automatic load balancing and scheduling systems, scientific visualization and virtual reality environments, data analysis and data retrieval systems (e.g. for BioInformatics), web based computing, etc. Problem Solving Environments : Tailored towards specific application fields (e.g. climate models) or generic environments for e.g. interactive simulation. The main challenge for computational science within ASCI lies in obtaining a true fundamental understanding of all issues related to the modeling and simulation chain. This can be achieved by thorough study of generic models and execution environments, and through detailed analysis of (possibly application oriented) case studies, in combination with realization of specific tools and problem solving environments.

B: Large-scale Distributed Information Systems and Embedded Systems With the advent of the Internet, a new dimension has been added to the notion of computing. Computing not only concerns high-performance computations, but includes collecting, processing, and communicating information across large-scale networks. These developments are leading to a next generation of distributed computer systems that are characterized by their mixed scale, interconnectivity, and heterogeneity. Examples include traffic control systems and large-scale computational grids. The number of embedded system applications is growing explosively and parallels the rapidly emerging systems-on-silicon paradigm and the explosive expansion of networked applications, both wired and wireless. Considerable research is needed by multidisciplinary teams of experts. Research topics include the following: B1: Operating systems Simply using existing operating systems to build the type of applications mentioned above will not do. Existing operating systems are tailored for general use and often require significant CPU power and main memory. In addition, they are still optimized for using local resources, and provide only traditional networking support. What is often needed are specialized operating systems that exploit the available networking facilities such as programmable network interfaces. In addition, the trend to customize operating systems to specific applications requires that new avenues to flexibility are explored. B2: Distributed Systems An important area of research covers middleware, also known as distributed systems. Middleware provides a layer of abstraction by which the underlying network infrastructure is mostly transparent to applications. Such a layer handles issues such as failure masking, resource management, caching and replication, automated storage, security, and so on. It provides applications with a convenient communication interface while hiding the intricacies related to issues such as performance and reliability. Developing scalable distributed systems is a major multifaceted research topic and includes subjects such as adaptive large-scale replication, worldwide directory services, and high-performance communication infrastructures. B3: Embedded systems An embedded system is a (not self-contained) part of a larger unit that provides service(s) to, or controls that unit. It typically consists of a heterogeneous collection of autonomous subsystems (ASICs, FPGAs, microcontrollers and DSPs) that require co-operation to perform (complex) tasks. They can be part of a geographically distributed system, locally distributed systems (in-home networks), or lumped systems (such as TV receivers). Reactive real-time embedded systems are becoming ubiquitous. Such systems react continuously to their environment at the speed of the environment. Research in this area ranges from the development of hardware and software components for digital signal processing to the development of advanced communication software. Formal methods must be called for to guarantee time-to-market, correctness and safety. However formal specifications, verifications and synthesis of software and hardware from high levels of abstraction have been demonstrated only for small, specialized languages with restricted semantics. This is at odds with the complexity and heterogeneity found in typical embedded systems of tomorrow. One challenge is to base the approach of designing heterogeneous reactive embedded systems on the use of appropriate formal models to describe the behavior of the system at a high level of abstraction, before a decision on its decomposition into hardware and software components is taken. Another challenge is the systematic development of embedded applications such as networked embedded applications. In the same way as embedded system design is a hardware/software co-design effort, embedded application and embedded system should be specified and explored in concert. Important issues here are scalability, robustness, re-use, quality of service, adaptive resource contracts, balancing processing load and data allocation over available processing and storage resources, both local and distributed.


12

C: Image and Multimedia Sensing, Processing, Interpretation, and Visualization On the junction of Imaging and Methods and Algorithms we deal with the development and evaluation of generic models, methods, algorithms and tools for image sensing, processing, interpretation, and visualization that can be used in many concrete applications. The past decade has witnessed a huge increase in the amount of digital visual data that is generated, processed, and stored. Because humans are good in the processing of visual information, and computers are strong in the fast processing of huge amounts of data, the best of both worlds can only be combined if computers can adequately process visual data for analysis and visualisation. Many theoretical and technological breakthroughs are required before we reach that goal. New fundamental insight into the inherent complexity of problems, new algorithms with proven practical behaviour are required. Research topics include: C1: Image and multimedia sensing The sensing of the real world with satellites, cameras, all kinds of (medical) scanners etc. The resulting data streams are typically large amounts of digital still images and video. Important aspects of research are the development of technology for image acquisition, image sensors and other media, motion capturing, image filtering, image coding, restoration, high quality image compression, and conversion. C2: Image processing Taking an image as input, process the digital image data, and output data either as an image, or as geometrical, numerical, or statistical data. Fundamentally difficult issues are topics such as: representation of images, segmentation, digital geometry, measurement, and mathematical morphology. C3: Image interpretation Classification and recognition of information in images. The problem is often ill-posed, and the image data often ambiguous. Challenging research questions are the development of new models for sensor data fusion, parameter estimation, adaptive control, statistical and geometrical pattern recognition, neuro computing, learning, goal directed computation, model-based interpretation, image database techniques. C4: Visualization and modeling The (interactive) modeling of complex objects and whole scenes, and the generation of images from these descriptions. Relevant research aspects are multi-modal data integration, 3D modelling and feature recognition, motion and path planning, computational geometry, data and information visualisation, rendering, 3D interaction, virtual environments, levels of detail. This list is not complete, and the classification of topics is not strict, there is much overlap between various topics. Each of the techniques plays a natural role in many applications, including bio-medical imaging, industrial inspection, image and video archiving and retrieval, document analysis, geographic information systems, earth observation, computer-aided design and reverse engineering, robotics and manufacturing, robot vision and path planning. See also multimedia information systems at the junction of Imaging and Communication and Systems. D: Multimedia Information Systems Multimedia systems are systems that are concerned with the handling of multimedia. It includes the acquisition, representation, composition, interaction, architecture, analysis, retrieval and distribution of one or multiple information streams like images, video, speech, sound, free text, language, and graphics. New uses of information require new ways of representing the information, not only because of the availability of multiple information streams but also because of their interaction and the way they can support one another. New methods, techniques and means for the manipulation and architecture of the information are required. It will not only influence the user and the use of information, extensive use of multimedia will also fire back at the internals of the computer, the computer organization and the networks connecting them. Research topics include: D1: Multimedia repositories The processing of multimedia information relies on new designs of the system architectures. Research questions are directed towards data and storage structures for multimedia data. Efficient use of storage space is necessary. Indexing multimedia data based on a higher, more abstract level are to be investigated. New search techniques to handle the large databases efficiently and to uncover information are to be developed. D2: Image, video and audio compression Although substantial progress has been achieved in efficient representation (compression) of images and video, it is no longer studied as an isolated problem, but is considered as one of the components of an entire system. Compression

Research Themes


optimization is now investigated in relation with overall performance, including limited or adaptive bandwidth, quality of service, scaling of information and the ability to analyze and to manipulate content in the compressed domain directly. To interact with multimedia data, techniques are investigated to describe the data as coherent spatio-temporal objects. D3: Multimedia understanding At the multimedia level, understanding the information is reached by fusing the information of different modalities and sources. The main issue is how to combine multiple information streams to facilitate a better understanding. Learning from multimedia repositories is a challenging problem. Multimedia mining covers the conversion of the content of large databases into generic rules on the content. D4: Multimedia editing systems For the design of multimedia products the availability of multimedia editing systems is indispensable. Research questions are how to make these systems work on a sufficient level of abstraction in order not to obstruct the designer, but which are - at the same time - intuitive to understand. D5: Delivery To secure multimedia data through e.g. the Internet, watermarking, security and encryption techniques are inevitable for copy protection, authentication or hiding of data. Research issues are for instance the development of watermarking techniques that are resistant to various deformations and privacy guaranteed exchange of multimedia data. D6: Multi-modal man-machine interaction The need grows for personalized video delivery, that is, for developing systems able to filter the incoming programs, to analyze them and prepare them for retrieval according to user preferences. This is only possible if such systems are capable of learning about the user profile. An important research issue here is to find ways of characterizing, quantifying and measuring emotions. D7:Multimedia applications Multimedia applications are found in many different circumstances. To mention a few examples, disciplines such as the textile branch, detective work and the medical profession drive on multimedia files. The management of data in bio-informatics, especially in the area of bio-diversity and microscopic analysis and the spatial location of metabolic processes invariably are multimedia of nature. Exploring the depth of the content is an important task for the development of new knowledge. In the Internet based tele-working and other aspects of professional life, multimedia transport of information to escape human transportation is of great societal relevance.

Chapter 3: ASCI –wide Events

14

Chapter 2 Scientific Output along Research Themes

For each ASCI subtheme this chapter contains a subsection in which the contribution of the various research groups in ASCI has been collected. The following scheme lists the enrollment of the groups with respect to the subthemes (x.y.z in a cell (group, theme) refers to the section in which the scientific ouput of group with respect to theme is reported).

A B C D VU-WI-I 2.2.6 UvA-FdNWI-caps 2.2.1 UvA-FdNWI-scs 2.1.4 UvA-FdNWI-ias 2.3.8 UvA-FdNWI-isis 2.3.13 2.4.2 UvA-FdNWI-csa 2.2.9 TUD-EWI-st-pds 2.2.7 TUD-EWI-me-ce 2.1.3 2.2.8 TUD-EWI-mm-cgcc 2.3.12 TUD-EWI-mm-ict 2.3.4 2.4.1 TUD-TNW-tn-qi 2.1.1 2.3.6 UL-WI-I(LIACS) 2.1.2 2.2.2 2.3.7 UL-LUMC-lkeb 2.3.14 UU-WI-ics 2.3.10 TUD-L&R-frs 2.3.11 UG-CS-is 2.3.15 UG-CS-svcg 2.3.1 TUE-EE-dmes 2.2.3 2.3.3 TUE-WI-vis 2.3.2 TUE-BMT-bmia 2.3.5 UT-EEMCS-dacs 2.2.4 EUR-RMI-bigr 2.3.9

2.1 A: High Performance Computing and Computational Science

2.1.1 Contribution of TUD-TNW-tn-qi Architectures and Algorithms, Embedded Imaging Systems The work in 2005 encompassed projects on embedded systems that incorporate real-time computer vision subsystems. The theoretical work was on the establishment of a parallel programming environment where heterogeneous architectures can be programmed within a single parallel programming language but use various programming paradigms, such as dataflow and data parallel. This is performed within the context of the Progress project SMARTCAM; with TUE, Philips Research and Philips CFT as main industrial partners. Data-parallelism is hidden in “algorithmic skeletons, while task parallelism is scheduled at run time over a Smartcam processor configuration incorporating TriMedia, Xetal-SIMD, FPGA with embedded PowerPC; and controller PC. Hardware design space exploration is part of the project. A real-time stereo vision camera was developed in cooperation with Philips Research and Philips Applied Technologies. Equivalent problems, navigation using cameras and encoders, are encountered in augmented reality applications. Within the Mobile & Wireless project of the Telematica Institute and with TUD-ITS, the positioning of virtual objects exactly in overlay with the real world, is obtained by determining the position and orientation of the human head in relation with the real world. An embedded imaging system closely coupled in a fast loop with the graphics animation system determines from visual cues, fused with information from gyros, accelerometers, magnetometers and tilt sensors, the position and orientation in 3D space with an update rate of 10ms. A demonstrator helmet was developed. A collaboration project with the Royal Academy of Art in the Hague to require Augmented Reality content was successfully applied for (start in 2006). A more or less equivalent problem can be found in the "ROBOCUP" project, in which two teams of four autonomous robots play soccer on an indoor field of about 10x6 meter. Positioning of the robot, interception of the ball, and collision avoidance of the other robots, is performed using odometry and real time image sensing and processing. This project was done in cooperation with the TU-Eindhoven. Similarly robot soccer with Sony AIBO dogs is played in cooperation with the UvA, UU and the Decis lab. A collaboration regarding humanoid robots is started up with the faculty of 3ME at the TU-Delft. A focus in these projects is on reinforcement learning of robots to walk and dribble with the ball.


Much effort was put into the set-up of the Dutch Initiative for Cure and Care Technology, a 3TU- initiative together with Erasmus MC, AMC and LUMC. Its focus is on the effectiveness of the entire cure and care chain as well as on automatic systems that sense-think-act. External projects SmartCam Period: 2002-2006; funding STW-Progress, with: TU/e Embedded Systems Papers in international journals Han, J., J. Gao, Y. Qi, P.P. Jonker, and J.A.B. Fortes; Toward Hardware-Redundant, Fault-Tolerant Logic for Nanoelectronics, IEEE Design & Test of Computers, vol. 22, no. 4, 2005, 328-339. Jonker, P.P.; Discrete topology on N-dimensional square tessellated grids, Image and Vision Computing, vol. 23, no. 2, 2005, Feb., 213-225. Contributions to international conference proceedings Broers, H., W. Caarls, P.P. Jonker, and R. Kleihorst; Architecture Study for Smart Cameras, Proc. EOS Conference on Industrial Imaging and Machine Vision (Munich, Germany, June 13-15), European Optical Society, 2005, 39-49 Fatemi, H., H. Corporaal, T. Basten, R. Kleihorst, and P.P. Jonker; Designing Area and Performance Constrained SIMD/VLIW Image Processing Architectures, in: J. Blanc-Talon, W. Philips, D. Popescu, P. Scheunders (eds.), Advanced Concepts for Intelligent Vision Systems - ACIVS 2005 (Proc. 7th Int. Conf., Antwerp, Belgium, Sep.20-23), Lecture Notes in Computer Science, vol. 3708, Springer Verlag, Heidelberg, 2005, 689-696. Fatemi, H., H. Corporaal, T. Basten, R. Kleihorst, and P.P. Jonker; Parallelism Support in SIMD/VLIW Image Processing Architectures, in: B.J.A. Krose, H.J. Bos, E.A. Hendriks, J.W.J. Heijnsdijk (eds.), ASCI 2005; Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging (Heijen, NL, June 8-10), ASCI, Delft, 2005, 291-296. Jonker, P.P., and H. Corporaal; Skeletons and Asynchronous RPC for Embedded Data- and Task Parallel Image Processing, Proc. 9th IAPR Conference on Machine Vision Applications (Tsukuba Science City, May 16-18), Tokyo, 2005, 384-387. Jonker, P.P., B. van Driel, J. Kuznetsov, and B. Terwijn; Algorithmic foundation of the Clockwork Orange Robot Soccer Team, in: M.Erdmann, D. Hsu, M. Overmars, A.F. van der Stappen (eds.), Algorithmic Foundations of Robotics VI, STAR 17 (Proc. 6th Int. Workshop, Zeist/Utrecht, July 11-13 2004), Springer Tracts in Advanced Robotics, vol. 17, Springer Verlag, Heidelberg, 2005 (@2004), 17-26. Schuitema, E., D.G.E. Hobbelen, P.P. Jonker, M. Wisse, and J.G.D. Karssen; Using a controller based on reinforcement learning for a passive dynamic walking robot, Humanoids2005, Proc. IEEE-RAS Int. Conf. on Humanoid Robots (Tsukuba, Japan, Dec.5-7), IEEE, 2005, 1-6. Cooperations within ASCI TUE-EE-dmes

2.1.2 Contribution of UL-LIACS Analysis and modeling of the circadian pacemaker at the cellular level The project for developing computer models and performing large-scale simulation of the circadian pacemaker at the cellular level started in 2004 as collaboration between the Leiden Institute of Advanced Computer Science (LIACS) and the Leiden University Medical Center (LUMC). In 2005 a simulation model has been developed to investigate the effects of the single unit activities (SUAs) on the activity of populations in the biological clock, more precisely the Suprachiasmatic Nuclei (SCN)). The simulations use a limited number of neurons and contain different light-dark regimes. Performance predictions and resource management in a Grid To deliver nontrivial quality of service is one of the most challenging problems in the Grid. Specifically, mining performance data on Grid resources to extract useful information, helps to improve performance or manage the system itself. The main research questions are: 1. How to define a set of performance metrics and efficiently predict it using statistical data mining techniques? 2. How resource brokers can incorporate prediction information in an intelligent way to improve throughput and response time on a meta-level? 3. How dynamic prediction information can be efficiently published or routed to the interested brokers/parties?


16

Another research topic is workload characterization and modeling in a Grid environment, specifically in Markov Arrival Processes and Markov Modulated Poisson Processes (MMPP) in modeling job arrival patterns at the Grid level. Cyttron Bio-Computing Search Project The Cyttron (www.cyttron.nl) consortium wants to implement a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail. We would like to provide a generic tool for identifying the molecular causes of disease, essential for the prevention of disease and the development of new drug and therapies, and to establish a platform for advanced diagnosis and tuning of individualized therapy, increasing effectiveness in health care. The consortium is highly multidisciplinary including (bio-)physicists, chemists, mathematicians, bio-informatics and image processing specialists, cell biologists, microscopists and medical researches from various research institutes. This sub-project of Cyttron focuses on image search algorithms and methods for bio-image data bases. HIRLAM on a grid environment We investigate how the operational numerical weather forecast system HIRLAM could be made grid-enabled, meaning what kind of adaptions are required to execute the system efficiently on a grid. As grid platform we use the DAS-2 system, which consists of five clusters located on five different universities. For the experiments on a single cluster, we found that more compute processors result in a better performance. However, we cannot obtain more improvement by only increasing the number of computer processors due to communication overhead. We also found that an additional one or two IO processors is an ideal number for the forecast model, more IO processors do not result in a better performance. For the experiment on a multi-cluster, we varied the distribution of processors over two clusters to investigate the performance of the forecast model. The distribution over two clusters increases the elapsed time: the more processors are distributed over the two clusters the larger the elapsed time becomes. Our experiment shows that running the forecast model on a multi-cluster does not result in a better performance than on a single cluster. However, using grid technology may give an advantage if the in- and output files are not located on the main compute server. Iterative Compilation In iterative compilation we search for the best program transformations by profiling many variants and selecting the one with the shortest execution time. Since this approach is extremely time consuming one has to incorporate static models. We show that a highly accurate model as a filter to profiling can reduce the number of executions by 50%. We also show that using a simple model to rank transformations and profiling only those with highest ranking can reduce the number of executions even further, in case we have a limited number of profiles at our disposal. We conclude that a production compiler might perform best using the last approach. Exterrnal Projects Cyttron Bio-Computing Search Project 2004-2008, Bsik, 8.8 MEuro. The Cyttron consortium wants to implement a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail. We would like to provide a generic tool for identifying the molecular causes of disease, essential for the prevention of disease and the development of new drug and therapies, and to establish a platform for advanced diagnosis and tuning of individualized therapy, increasing effectiveness in health care. Distributed ASCI Supercomputer 3 (DAS 3) 2005-2009, NWO-EW, 900 KEuro, cooperation with all partners of the ASCI graduate school. Flexible Application Mapping Environments (FAME) 2002-2006, NWO-EW, 270 KEuro, TUE The FAME project aims at developing a novel approach to program optimization, namely, iterative compilation in which the transformation space is searched and profiling is used to measure the impact of transformations. One of the most important goals of this project is to develop heuristics to control the complexity of the search by using analytical models and domain specific knowledge. Lateral boundary conditions for nested models – mathematical and computational aspects 2005-2009, NWO-ALW/EW, 300 KEuro. The aim of this project is to formulate and implement good lateral boundary conditions, both mathematically ("well-posed") and meteorologically (i.e., properly transferring information from the global into the local domain). The reseach should result in a production code, implying that the code implements the developed algorithms efficiently on massively parallel and/or GRID computers. Papers in international journals D. Cheresiz, B.H.H. Juurlink, S. Vassiliadis, and H.A.G. Wijshoff, The CSI Multimedia Architecture, IEEE Trans. VLSI Syst. 13(1):1-13 (2005).


Contributions to international conference proceedings F.J. Cazorla, P.M.W. Knijnenburg, R. Sakellariou, E. Fernandez, A. Ramirez, and M. Valero, Architectural Support for Real-Time Task Scheduling in SMT Processors, proceedings CASES, pp. 166-176, 2005. F.J. Cazorla, P.M.W. Knijnenburg, R. Sakellariou, E. Fernandez, A. Ramirez, and M. Valero, Quality of service for Simultaneous Multithreading Processors, proceedings Advanced Computer Architecture and Compilation for Embedded Systems (ACASES), Poster Abstracts, pp. 67-70, 2005. M. Haneda, P.M.W. Knijnenburg, and H.A.G. Wijshoff, Automatic Selection of Compiler Options using Non-Parametric Inferential Statistics, proceedings PACT, pp. 123-132, 2005. M. Haneda, P.M.W. Knijnenburg, and H.A.G. Wijshoff, Generating New General Compiler Optimization Settings, proceedings ICS, pp. 161-168, 2005. M. Haneda, P.M.W. Knijnenburg, and H.A.G. Wijshoff, Optimizing General Purpose Compiler Optimization, proceedings Computing Frontiers, pp. 180-188, 2005. Hui Li, David Groep, and Lex Wolters, Efficient Response Time Predictions by Exploiting Application and Resource State Similarities, in proceedings of the 6th IEEE/ACM International Workshop on Grid Computing (Grid2005), Seattle, Washington, USA, November 2005. Hui Li, David Groep, and Lex Wolters, Workload Characteristics of a Multi-cluster Supercomputer, in D. Feitelson, L. Rudolph, and U. Schiegelshohn (Eds.): proceedings of the 10th Workshop on Job Scheduling Strategies for Parallel Processing (JSSPP 2004), LNCS 3277, pp. 176-193, 2005. Experimental software H. Li, Performance Data Miner (PDM): A Java Toolkit for Mining Performance Data in the Grid(www.liacs.nl/~hli/pdm/). PDM is a Java-based toolkit for mining the performance data in distributed computing environments, or Grids. Its goal is to extract useful information which can be used to improve performance or manage the Grid system itself. Experimental hardware DAS-2 & DAS-3 Cooperations within ASCI With research groups of prof.dr. H. Corporaal/ TUE, dr. D. Epema/ TUD, prof.dr. S. Vassiliadis/TUD

2.1.3 Contribution of TUD-EWI-me-ce ∆-iliad The ∆-ILIAD research concerns with new computer architectural paradigms. The gamma of processor architectures considered include general purpose, domain (e.g. media), vector processor extensions, polymorphic processing and non-conventional architectures. Furthermore, we perform some reality checks for existing processor implementations. More specifically the Delta-Iliad team is currently working on the following research topics: Vector Facility Traditionally, vector processors are limited by memory accesses, sectioning, and simple-minded computations. The ∆-ILIAD vector architecture eliminates sectioning, alleviates storage access overhead by overlapping accesses with computations and merging both of them into a single instruction. In addition to traditional operations, ∆-ILIAD architecture includes new instructions that perform complex multicycle latency operations. With the introduction of the ∆-ILIAD mechanism a substantial code elimination is achieved. The specific dense and sparse architectural mechanisms of ∆-ILIAD include the Complex Streamed Instruction Set (CSI) and the ∆-ILIAD sparse vector processing, which are described below. CSI Media Architecture The Complex Streamed Instruction Set Architecture (CSI) is a memory-to-memory vector architecture targeted at multimedia applications. A single CSI instruction can process data streams of arbitrary length and, in addition to traditional arithmetic and logical operations, performs data accesses, conversion between storage and computation formats (packing and unpacking), and complex arithmetic hardwired computation. The main new features of the CSI are elimination of the vector sectioning instructions, elimination of the packing/unpacking instructions, and introduction of new complex media related arithmetic instructions.


18

Sparse Matrix Architectures Vector processors are known for performing good on large amounts of regular data. However, when operating on sparse matrices such as the one depicted here, the irregular structure induces a performance degradation. The main reasons are the need for expensive indexed memory accesses and high vector startup overhead due to short vectors. Moreover, the need for positional information when storing sparse matrices implies an extra storage overhead. The aim of this project is to aleviate most of the aforementioned problems and increase the efficiency of vector processors on sparse operations. This is achieved by introducing a new block based Sparse Martix format. In conjunction with a harware Vector ISA extension and specialized hardware for sparse matrix computations we can aleviate the need for indexed memory accesses. Speedups of 4-5 times have been obtained for matrix-vector multiplication, an important kernel in sparse matrix processing. Delft Sparse Architectures Benchmark The Delft Sparse Architecture Benchmark (D-SAB) has been developed in the Computer Engineering Laboratory as a part of the ∆-iliad project. Its purpose is the evaluation of novel architectures and techniques for processing Sparse Matrices. The benchmark comprises two parts: The benchmark operations and the benchmark matrices. Polymorphic processors Current processor architectures force a complete separation of tasks between implementations (hardware-architectures), which interpret an architecture, and the programmer targeting this architecture. Polymorphic processors eliminate the gap between the (hardware) implementations and the programmer of the hardware. This is achieved using a new programming paradigm and emulation on reconfigurable hardware. Delft Linpack The TOP500 Supercomputer Sites webpage ( http://www.top500.org/) presents the world best highperformance computers. The LINPACK Benchmark is used as a yardstick for performance. Companies like IBM, HP, NEC and Intel (ASCI red) are presented there with their top supercomputers. The interesting question is: Can a university student team with out of the shelf inexspensive hardware components beat the industry supercomputers on Linpack? The DelftLinpack-1 processor uses the power of reconfigurable hardware in order to attempt an answer of this question. Xilinx state of the art FPGA (XC2VP50) incorporating reconfigurable logic and four PowerPC general purpose cores will be used to implement the DelftLinpack-1 machine. Contributions to Books D. Cheresiz, B.H.H. Juurlink, S. Vassiliadis, H. A. G. Wijshoff, The CSI Multimedia Architecture, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, pp. 1-13, January 2005, Vol. 13, No. 1 Contributions to international conference proceedings S. Vassiliadis, L. A. Sousa, G. N. Gaydadjiev, The Midlifekicker Microarchitecture Evaluation Metric, Proceedings of the IEEE International conference on Application-Specific Systems, Architectures and Processors (ASAP05), pp. 92-97, Samos, Greece, July 2005 B.H.H. Juurlink, A. Shahbahrami, S. Vassiliadis, Avoiding Data Conversions in Embedded Media Processors, Proceedings of the 20th ACM Symposium on Applied Computing, pp. 901-902, Santa Fe, New Mexico, USA, March 2005 S. Suijkerbuijk, B.H.H. Juurlink, Implementing Hardware Multithreading in a VLIW Processor, Proc. 17th Int. Conf. on Parallel and Distributed Computing and Systems, pp. 674-679, Phoenix, AZ, USA, November 2005

2.1.4 Contribution of UvA FdNWI-scs Modelling and Simulation Problem Solving Environments and Visualisation Within each of the two main themes in our research, a number of projects are being pursued. Sometimes these projects intersect both themes. We will start this section with the one thesis defended in 2005, by K.A. Iskra, and then move on to a thematic report on the projects that were active in 2005. Besides the many projects that were active in 2005, a large number of proposals were written in 2005 and granted in 2005 or early 2006. These include the EU projects Virolab, ACGT, and QosCos, a project on Computational Finance an NWO STARE project.. Much of our research is performed in close international collaborations. Within the framework of a formal agreement between the Universiteit van Amsterdam and the Saint Petersburg State University, Russian Federation, a joint laboratory for computational science was established in Saint Petersburg. This lab works on problem solving environments and applications. We formally collaborate with CYFRONET, Cracow, Poland, through joint Ph.D. students in the field of monitoring and scheduling for scientific computing on the Grid. In the framework of a recently started NATO Science for Peace program we collaborate with institutes of the Siberian Branch of the Russian Academy of Science in the field of biomedical diagnostic systems. Finally, strong research cooperation is established with a number of USA based universities.


PhD Thesis Kamil Iskra defended his thesis “Time Warp – from Cluster to Grid” in June 2005. In this thesis he investigates the behaviour of the Time Warp optimistic parallel discrete event simulation strategy in high communication latency environments. Various methods to reduce the impact of the high latency are studied. Within each of the two main themes in our research, a number of projects are being pursued. Sometimes these projects intersect both themes. We will start this section with the one thesis defended in 2005, by K.A. Iskra, and then move on to a thematic report on the projects that were active in 2005. Besides the many projects that were active in 2005, a large number of proposals were written in 2005 and honoured in 2005 or early 2006. External Projects Modelling and Simulation AmCG People involved J.A. Kaandorp, J. Cui, P.M.A. Sloot.. The bioinformatics group of Amsterdam Genomics Center (AmGC) is a collaboration between researchers from SILS-UVA, IvI-UvA and Academic Medical Center Amsterdam working on bioinformatics. The bioinformatics group is coordinated by Antoine H.C. van Kampen (Academic Medical Center Amsterdam) and Jaap Kaandorp (Section Computational Science, UvA). One of the major current research themes within the AmGC is the analysis and modelling of biological networks. The bioinformatics of biological networks involves a broad range of research and approaches. This research includes topics like identification of common regulatory elements for genes in a pathway, the modelling and simulation of pathways, the reconstruction of pathways from experimental data, the visualization of pathways, and the representation of pathways in graphs and databases. To accelerate our understanding of the (dynamics of) biological networks, it is imperative that these efforts are combined. Subsequently they have to be applied to real biological problems. In 2004 three projects related to modelling and simulation of biological networks have been started up within the Section Computational Science (Mesoscale simulation paradigms for biological systems, Simulation of developmental regulatory networks, Mathematics and Computation for the System Biology of Cells). In May 2005 the bioinformatics group organised the second international symposium on networks in bioinformatics in Amsterdam. In November 2005 the project Modelling and inferring developmental regulatory networks was started up. Mesoscale simulation paradigms for biological systems People involved J.A. Kaandorp, J. Cui. Project funded by the Applied Mathematics programme of the Dutch Science foundation, duration 2003 - 2007, total value Eur 329000 In this proposed project we want to develop and compare computational models of parts of the living cell that can calculate in detail system properties from experimentally obtained molecular and physical-chemical data. Such a model is as close as possible to the biological experiments and therefore can be used not only for understanding the principles of function but also to steer further biological experiments. Simulation of developmental regulatory networks People involved J.A. Kaandorp, Y. Fomekong Nanfack Funded by Computational Life Sciences programme of the Dutch Science foundation, duration 2003 - 2007, total value Eur 328000 In this project we will develop a model for simulating regulatory networks that are capable of quantitatively reproducing spatial and temporal expression patterns in developmental processes. The model is a generalization of the standard connectionist model used for modelling genetic interactions. The model will be coupled with a biomechanical model of cell aggregates and used to study the formation of spatial and temporal expression patterns of gene products during development in cellular systems (sponges and scleractinian corals). Mathematics and Computation for the System Biology of Cells (Cell.Math.) People involved J.A. Kaandorp, J. Vidal Rodriguez, P.M.A. Sloot. Project funded by Computational Life Sciences programme of the Dutch Science foundation, duration 2003 - 2007, total value Eur 487000 The aim of the project is to develop, implement, and validate mathematical and computational techniques for the systems biology of the cell. Biologists and mathematicians together will formulate realistic mathematical models of metabolic and regulatory networks including intrinsic spatial non-homogeneity. Depending on the cellular phenomenon considered, models and methods of appropriate temporal and spatial scales will be developed and can then be applied: models in the form of ordinary differential equations and methods for system reduction; multi-adaptive computational methods for partial differential equations (PDEs) for moderate spatial and temporal variability within a cell or an organelle; particle models describing the interaction of individual molecules and computational methods for the evaluation of the dynamic behaviour; and methods for integration of these different approaches into a single simulation.


20

Modelling and inferring developmental regulatory networks People: J.A. Kaandorp, and Y. Fomekong Nanfack, P.M.A. Sloot. Project funded by the Dutch Russian Research Cooperation 2004 (interdisciplinary mathematics), Total value Eur 123000 The project is a collaboration with Prof. A. Samsonov (Ioffe Institute Russian Academy of Sciences, St Petersburg). And Prof. M. Samsonava (St Petersburg State Polytechnical University) In this project we will develop mathematical models of regulatory networks that are capable of quantitatively reproducing spatial and temporal expression patterns in developmental processes. The second aim of the project is to develop new statistical techniques for the analysis of temporal and spatial gene expression patterns, model driven optimization methods, based on simulated annealing in combination with semi parametric statistical techniques, to infer regulatory networks from actual data sets. To demonstrate the effectiveness of our models we are planning to apply these methods to analyze the developmental process in organisms with a relatively simple body plan (for example sponges and scleractinian corals). Virolab People: P.M.A. Sloot et al. Funding EU, 3.5 MEuro, 1.0 MEuro (UvA) Funding was obtained for Virolab in 2005; the actual project will start in 2006. In Virolab, the SCS group will extend their work on the modelling of HIV infections started in the BMI project. Nato SfP People: A.G. Hoekstra, M Yurkin, K. Gilev, K. Semianov In the context of the Science for Peace program of Nato we collaborate with institutes in Minsk (Belarus) and Novosibirsk (Russian Federation) on new sensitive methods for cytological analysis of haematological samples. We concentrate on the computational science aspects and HPC simulations in the field of computational electromagnetics of the project. We delivered ADDA, a public domain code for Computational Light Scattering, capable to simulate scattering from particles as large as real biological particles. In effect, this code is capable to solve sets of linear equations with O(107) unknowns. ADDA is a highly efficient parallel code that was ported to many parallel computers, including the USA ASCI computers, and the Dutch Asci Supercomputer DAS. LBM A.G. Hoekstra, L. Abrahamyan, P.M.A. Sloot The work on mesoscopic modelling and simulation concentrated on adapting the LBM method, and specifically the L-BGK method, for unsteady flow, and to apply this to modelling flow of blood during a full heart beat in the lower abdominal aorta. We also initiated research to adapting the models to take advantage of state of the art (parallel) numerical algorithms (such as multi grids). Moreover we studied fluid-structure interaction in LBM, and have demonstrated that for 2D time-harmonic flow coupled to a simple elastic wall a theoretical expression for a dispersion relation is recovered, thus demonstrating the correctness of our model. In collaboration with Prof. Reiber of LUMC we have developed a problem-solving environment for image-based computational haemodynamics. Moreover, we have studied in detail the stability of time harmonic LBM simulations and addressed the question of optimally choosing simulations parameters, by targeting at a specified wanted accuracy and minimizing execution time. Finally, in collaboration with a number of German institutes, we have developed a next generation LBM code capable of handling nested grids. We have realised a state-of-the-art decomposition tool to aid parallelization of this unstructured LBM code. Part of this research is supported through the Token 2000 project ‘DIME’. PECVD Involved: V.K. Krzhizhanorskaya, P.M.A. Sloot In this project a grid-based PSE to study the plasma enhanced chemical vapour deposition of thin films is being developed. Internally funded project in close collaboration with the IHPCIS in St. Petersburg. The research was conducted with financial support from the Dutch National Science Foundation NWO and the Russian Foundation for Basic Research under grants number 047.016.007 and 047.016.018, and with partial support from the CrossGrid EU project IST-2001-32243. Cluster of Grapes People involved: S. Portegies Zwart, A. Gualandris, M.S. Sipior, E. Gaburov, B. Bastijns, P.M.A. Sloot, G.D. van Albada + collaboration with the Pannekoek Astronomical Institute One of the other research areas in which the SCS group is active, is computational astronomy. "A Cluster of Grapes", a joint NWO proposal with the astronomical institute "Anton Pannekoek" to NWO was honoured in 2001. The research covers a wide range of subjects on the boundary of Astronomy and Computational Science. As an example, Gualandris, Tirado-Ramos and Portegies Zwart studied the performance of a parallel astrophysical N-body solver on pan-European computational grids. It was demonstrated that especially for large problems grids constitute a suitable computational environment. Sipior is working to merge the “Kira integrator” with the GADGET tree code developed at MPI Garching. The end result will allow us to make use of the unique advantages of each numerical approach – the high precision of Kira for studying the dynamics of a large stellar cluster, and the computational speed of a tree code, used to simulate the galaxy in which the cluster is embedded. The first stage of this effort is now completed and undergoing testing.


Dynamical evolution of dense star clusters Portegies Zwart & McMillan considered the formation of the recently discovered ``hot Jupiter'' planet orbiting the primary component of the triple star system HD 188753. The current outer orbit of the triple is too tight for a Jupiter-like planet to have formed elsewhere and migrated to its current location, the binary may have been much wider in the past. We assume here that the planetary system formed in an open star cluster, the dynamical evolution of which subsequently led to changes in the system's orbital parameters and binary configuration. We calculate cross sections for various scenarios that could have led to the multiple system currently observed and conclude that component A of HD 188753 with its planet was most likely formed in isolation, to be swapped into a triple star system by a dynamical encounter in an open star cluster. We estimate that within 500 pc of the Sun, there are about 1200 planetary systems that, like HD 188753, have orbital parameters unfavourable for forming planets but still have a planet, making it quite possible that the HD 188753 system was indeed formed by a dynamical encounter in an open star cluster. A press release was sent out and is available at: http://www.uva.nl/onderzoek/ and articles about this work appeared in Noorderlicht, New Scientist and Der Spiegel. An animation of the proposed encounter can be viewed at http://modesta.scienc.uva.nl. Gualandris, Portegies Zwart and Sipior investigated the interplay between the supermassive black hole in the centre of the Milky-way Galaxy and its local stellar cluster. They simulations the interactions between the black hole and other stars to study the possibility that the hypervelocity star SDSS J090745.0+024507, found in the halo was initially ejected from this region. The star currently has a velocity exceeding 700 km/s moving away from the Galactic centre. With the measured radial velocity and the estimated distance to the star, they traced its trajectory backward in time in the Galactic potential. Assuming it was ejected from the centre, they found that a proper motion of about 2 milli-arcseconds /yr is necessary for the star to have come within a few parsecs of the SMBH. They conclude that this star is most likely ejected in a strong dynamical interaction between the supermassive black hole and a tight binary, of which the currently observed star once was a component. Such events should occur frequently enough that the Galactic halo should be populated with more than a hundred hyper velocity stars. Problem Solving Environments and Visualization CrossGrid People: A.G. Hoekstra, E.V. Zudilova, A. Tirado Ramos, R. Shulakov, D. Shamonin, P.M.A. Sloot, G.D. van Albada, M. Scarpa In the CrossGrid project, we participate in a large European collaboration of 21 partners in the development of interactive applications in a Grid environment. These environments are characterized by the participation of multiple, geographically distributed organizations, sharing computational resources and data. This gives an improved access to these resources and data, at a cost in network and scheduling delays. In CrossGrid, we provide an application aiming to support vascular surgeons in pre-operative planning. 2004 was a consolidation of a consortium-wide integration of grid support tools and applications. The result is a fully integrated virtual vascular surgery on the grid, which was successfully demonstrated during the European Grid Conference 2005 in Amsterdam. The CrossGrid project ended in March 2005, but its results will continue to play an important role in our future research efforts.

Shear stress, velocities,

masses, etc.

ce (CrossGrid)

se2 (D-VRE machine)

MD login and Grid certificates submission

Bypass creation LB mesh generation

Job submission Job monitoring

MR image Segmentation(soon a GS!)

se1 (e.g., Leiden)

Patient at MRI scanner

MR image

Virtual Node Creation

Virtual Operation Theatre

Simulated flow

ce (CrossGrid)

I have no GVK images,

ask Elena

Fig.1: CrossGrid


22

Token2000 People: A.G. Hoekstra, P.M.A. Sloot, E.V. Zudilova, M. Scarpa, L. Abrahamyan ASCI partner: LKEB/LUMC (P.I.: Reiber) Token2000 is a nationally funded project (NWO), where we collaborate closely with the Universities of Leiden and Twente on the development of an interactive medical application, somewhat similar to the work in CrossGrid. This application is intended for training of surgeons. In collaboration with Leiden University Medical Centre we have created Hemosolve, a problem-solving environment for image based computational Haemodynamics. Hemosolve includes our L-BGK solver, but also a FEM Navier-Stokes solver. Moreover, it contains a 3D editing tool and powerful visualization modules. High Performance Simulation on the GRID Dutch-Russian project: NWO-RFBS-047.016.007 People: P.M.A. Sloot, A.V. Bogdanov. In this project we study the use of Grid systems for High Performance Simulations. The project is a collaboration between Amsterdam, St. Petersburg, Moscow and Novosibirsk. Dynamite People: K.A. Iskra, T. Gubala, B. Ó Nualláin, D.A. Kaarsemaker, G.D. van Albada, P.M.A. Sloot. The Dynamite project (now internally funded) is the continuation of work on the dynamic scheduling and migration of tasks in parallel programs started in the ESPRIT project Dynamite. In 2005 work continued on the single task checkpointer (ironing out some smaller remaining defects) and on developing support for MPI and grid environments. PDES Involvement: K.A. Iskra, G.D. van Albada, P.M.A. Sloot. This research on parallel discrete event simulation is a continuation of the work by Overeinder on the behaviour of optimistic PDES, extended to grid-like environments. In 2005 further important results regarding the behaviour of PDES in a wide-area distributed environment were obtained. Important strategies like dedicated routers, message aggregation and lazy cancellation were studied and different methods of “global virtual time” evaluations were evaluated. The project culminated in the PhD defence by K.A. Iskra in June 2005. External Projects Projectname period Funding agency Total

funding partners in ASCI

Brief description

CrossGrid 2002 – 2006 EU, 5th framework k€ 405 Interactive Grid applications A Cluster of Grapes

2001 - 2005 NWO Computational Science

kf 690 N-Body simulations. P.I: E.van den Heuvel, UvA Astronomy

Mesoscale simulation paradigms for biological systems

2003-2007 NWO Applied Mathematics

k€ 329 See report

Simulation of developmental regulatory networks

2003-2007 NWO Computational Life Sciences

k€ 328 See report

Mathematics and Computation for the System Biology of Cells

2003-2007 NWO Computational Life Sciences

k€ 487 See report

Token 2000 2002 – 2007 NWO - EW k€ 150 LKEB/LUMC (P.I.: Reiber)

Mesoscopic bloodflow simulation

Doctoral Degrees K.A. Iskra: Time Warp - from Cluster to Grid, PhD thesis, University of Amsterdam, Amsterdam, The Netherlands, (Promotor: Prof. Dr. P.M.A. Sloot, Co-promotor: Dr. G.D. van Albada) June 2005. ISBN 90-5776-139-4, funded by University of Amsterdam,. Virtual Laboratory for e-Science project,. Netherlands Organisation for Scientific Research . Massively Parallel Programming programme (NWO.MPR),. European Union . ESPRIT project 23499 Dynamite. Contributions to Books Tirado-Ramos; D.P. Shamonin; R.M. Shulakov; D.J. Groen; A.G. Hoekstra; G.D. van Albada; E.V. Zudilova and P.M.A. Sloot: Interactive Problem Solving Environments on the Grid for Image-based Computational Haemodynamics, in The CrossGrid Book, (accepted 15 January 2004) The European CrossGrid Consortium, 2005.


Papers in international journals A.M.M. Artoli; A.G. Hoekstra and P.M.A. Sloot: Optimizing lattice Boltzmann simulations for unsteady flows, Computers & Fluids, vol. 35, nr 2 pp. 227-240. 2005. K.A. Iskra; G.D. van Albada and P.M.A. Sloot: Towards Grid-Aware Time Warp, Simulation: Transactions of The Society for Modeling and Simulation International, vol. 81, nr 4 pp. 293-306. April 2005. J.A. Kaandorp; P.M.A. Sloot; R.M.H. Merks; R.P.M. Bak; M.J.A. Vermeij and C. Maier: Morphogenesis of the branching reef coral Madracis mirabilis, Proc. Roy. Soc. B, vol. 272, nr 1559 pp. 127-133. January 2005. O.V. Kaluzhnaya; S.I. Belikov; H.C. Schröder; S. Zapf; A. Borejko; J.A. Kaandorp; A. Krasko; I.M. Müller and W.E.G. Müller: Dynamics of skeletal formation in the Lake Baikal sponge Lubomirskia baicalensis. Part I biological and biochemical studies, Naturwissenschaften, vol. 92, pp. 128-133. 2005. V.V. Krzhizhanovskaya; P.M.A. Sloot and Y.E. Gorbachev: Grid-based Simulation of Industrial Thin-Film Production, Simulation: Transactions of the Society for Modeling and Simulation International, (Special Issue on Applications of Parallel and Distributed Simulation) vol. 81, nr 1 pp. 77-85. January 2005.

K. Rycerz; M.T. Bubak; M. Malawski and P.M.A. Sloot: A Framework for HLA-Based Interactive Simulations on the Grid, Simulation: Transactions of The Society for Modeling and Simulation International, (Special Issue on Applications of Parallel and Distributed Simulation) vol. 81, nr 1 pp. 67-76. April 2005. P.M.A. Sloot; A.V. Boukhanovsky; W. Keulen; A. Tirado-Ramos and C.A. Boucher: A Grid-based HIV Expert System, Journal of Clinical Monitoring and Computing, vol. 19, nr 4-5 October 2005. ISSN: 11387-1307. J.M. Voogd; P.M.A. Sloot and R. van Dantzig: Equilibrium spherically curved two-dimensional Lennard-Jones systems , J. Chem. Phys., vol. 123, nr 084105 pp. 1-5. 2005. (DOI: 10.1063/1.2007707) M.A. Yurkin; K.A. Semyanov; P.A. Tarasov; A.V. Chernyshev; A.G. Hoekstra and V.P. Maltsev: Experimental and theoretical study of light scattering by individual mature red blood cells with scanning flow cytometry and discrete dipole approximation, Applied Optics, vol. 44, pp. 5249-5256. 2005. Z. Zhao; G.D. van Albada and P.M.A. Sloot: Agent-based flow control for HLA components, International journal of simulation transaction, special issue Agent Directed Simulation, vol. 81, nr 7 pp. 487-501. 2005. E.V. Zudilova and P.M.A. Sloot: Bringing Combined Interaction to a Problem Solving Environment for Vascular Reconstruction, Int. J. Future Generation Computer Systems, vol. 21, nr 7 pp. 1167-1176. 2005. A. Tirado-Ramos; D.P. Shamonin; R.M. Shulakov; D.J. Groen; A.G. Hoekstra; G.D. van Albada; E.V. Zudilova and P.M.A. Sloot: Interactive Problem Solving Environments on the Grid for Image-based Computational Haemodynamics, in The CrossGrid Book, (accepted 15 January 2004) The European CrossGrid Consortium, 2005. Contributions to international conference proceedings L. Abrahamyan; J.A. Schaap; A.G. Hoekstra; D.P. Shamonin; F.M.A. Box; R.J. van der Geest; J.H.C. Reiber and P.M.A. Sloot: A Problem Solving Environment for Image-Based Computational Hemodynamics, in V.S. Sunderam; G.D. van Albada; P.M.A. Sloot and J.J. Dongarra, editors, Computational Science - ICCS 2005: 5th International Conference, Atlanta, GA, USA, Proceedings, Part I, in series Lecture Notes in Computer Science, vol. 3514, pp. 287-294. Springer, Berlin, Heidelberg, May 2005. ISBN 3-540-26032-3.

A.G. Hoekstra and P.M.A. Sloot: Introducing Grid Speedup Gamma: A Scalability Metric for Parallel Applications on the Grid, in P.M.A. Sloot; A.G. Hoekstra; T. Priol; A. Reinefeld and M.T. Bubak, editors, Advances in Grid Computing - EGC 2005, in series Lecture Notes in Computer Science, vol. 3470, pp. 245-249. Springer, Berlin, Heidelberg, February 14-16 2005. ISBN 3-540-26918-5.

A.G. Hoekstra: Image-based computational hemodyanamics with the Lattice Boltzmann Method, in K.J. Bathe, editor, Computational Fluid and Solid Mechanics 2005, pp. 672-675. Elsevier Ltd, 2005. K. Kruszynski; R. van Liere and J.A. Kaandorp: Quantifying performance of 3D skeletonization algorithms, in IASTED International Conference on Visualization, Imaging, and Image Processing (VIIP 2005), Benidorm, Spain, September 2005. T.A. Kvaløy; E. Rongen; A. Tirado-Ramos and P.M.A. Sloot: Automatic Composition and Selection of Semantic Web Services, in P.M.A. Sloot; A.G. Hoekstra; T. Priol; A. Reinefeld and M.T. Bubak, editors, Advances in Grid Computing - EGC


24

2005, in series Lecture Notes in Computer Science, vol. 3470, pp. 184-192. Springer, Berlin, Heidelberg, February 14-16 2005. ISBN 3-540-26918-5.

K. Rycerz; M.T. Bubak; M. Malawski and P.M.A. Sloot: HLA Grid based support for simulation of vascular reconstruction, in S. Gorlatch and M. Danelutto, editors, TR-05-22, Proceedings of the CoreGRID Workshop "Integrated Research in Grid Computing", pp. 165-174. University of Pisa, Computer Science Department, Pisa, November 2005. P.M.A. Sloot; A.V. Boukhanovsky; W. Keulen and C.A. Boucher: A Grid-based HIV Expert System, in Proceedings of the 5th International Symposium on Cluster Computing and Grid, CCGrid, (IEEE product number: EX1055C) IEEE, Cardiff, Wales, UK, May 2005. ISBN: 0-7803-9075-X. P.M.A. Sloot; C.A. Boucher; M.T. Bubak; A.G. Hoekstra; P. Plaszczak; A. Posthumus; D. van de Vijver; S. Wesner and A. Tirado-Ramos: VIROLAB - A Virtual Laboratory for Decision Support in Viral Diseases Treatment, in Cracow Grid Workshop 2005, (in press, best poster award) Cracow, Poland, November 2005. A. Tirado-Ramos; D.J. Groen and P.M.A. Sloot: On-line Application Performance Monitoring of Blood Flow Simulation in Computational Grid Architectures, in The 18th IEEE Symposium on Computer-Based Medical Systems, special track Grids for Biomedicine and Bioinformatics, pp. 511-516. Trinity College Dublin, Ireland, June 23-24 2005. ISBN 0769523552. A. Tirado-Ramos; G. Tsouloupas; M.D. Dikaiakos and P.M.A. Sloot: Grid Resource Selection by Application Benchmarking: a Computational Haemodynamics Case Study, in V.S. Sunderam; G.D. van Albada; P.M.A. Sloot and J.J. Dongarra, editors, Computational Science - ICCS 2005: 5th International Conference, Atlanta, GA, USA, Proceedings, Part I, in series Lecture Notes in Computer Science, vol. 3514, pp. 534-543. Springer, Berlin, Heidelberg, May 2005. ISBN 3-540-26032-3. A. Tirado-Ramos; D.J. Groen and P.M.A. Sloot: Exploring OGSA Interoperability with LCG-based Production Grids for Biomedical Applications, in M.T. Bubak; M. Turala and K. Wiatr, editors, Cracow Grid Workshop 2005, pp. 367-374. Ekodruk, Cracow, Poland, November 2005. M.A. Yurkin; K.A. Semyanov; A.G. Hoekstra and V.P. Maltsev: Experimental and Theoretical Study of Light Scattering by Individual Mature Red Blood Cells with Scanning Flow Cytometry and Discrete Dipole Approximation, in Proceedings of the 8th Conference on Electromagnetic and Light Scattering by Nonspherical Particles: Theory, Measurement and Applications, pp. 333-336. Salobrena, Spain, May 2005. M.A. Yurkin and A.G. Hoekstra: Capabilities of the discrete dipole approximation for simulation light scattering by biological cells, in Proceedings of the NATO Advanced Research Workshop: Optics of Biological Particles, pp. 19-20. Novosibirsk,Russia, October 2005. M.A. Yurkin; D. de Kanter and A.G. Hoekstra: Applicability of effective medium approximation to light scattering by granulated biological cells - first results, in Proceedings of the NATO Advanced Research Workshop: Optics of Biological Particles, pp. 46-47. Novosibirsk,Russia, October 2005. Z. Zhao; G.D. van Albada and P.M.A. Sloot: Rapid Prototyping of Complex Interactive Simulation Systems, in Proceedings of the 10th IEEE International Conference on Engineering of Complex Computer Systems (ICECCS'05) , pp. 366-375. IEEE Computer Society Press, Shanghai, China, June 2005.

Z. Zhao; A.S.Z. Belloum; H. Yakali; P.M.A. Sloot and L.O. Hertzberger: Dynamic Workflow in a Grid Enabled Problem Solving Environment, in Proceedings of the 5th International Conference on Computer and Information Technology (CIT2005), pp. 339-345 . IEEE Computer Society Press, Shanghai, China, September 2005. Z. Zhao; A.S.Z. Belloum; A. Wibisono; F. Terpstra; P.T. de Boer; P.M.A. Sloot and L.O. Hertzberger: Scientific workflow management: between generality and applicability, in Proceedings of the International Workshop on Grid and Peer-to-Peer based Workflows in conjunction with the 5th International Conference on Quality Software, pp. 357-364. IEEE Computer Society Press, Melbourne, Australia , September 19th-21st 2005.

Z. Zhao; A.S.Z. Belloum; P.M.A. Sloot and L.O. Hertzberger: Agent Technology and Generic Workflow Management in an e-Science Environment, in Hai Zhuge and G.C. Fox, editors, Grid and Cooperative Computing - GCC 2005: 4th International Conference, Beijing, China, in series Lecture Notes in Computer Science, vol. 3795, pp. 480-485. Springer, November 2005. ISBN 3-540-30510-6. (DOI: 10.1007/11590354_61) Z. Zhao; A.S.Z. Belloum; P.M.A. Sloot and L.O. Hertzberger: Agent technology and scientific workflow management in an e-Science environment, in Proceedings of the 17th IEEE International conference on Tools with Artificial Intelligence (ICTAI05), pp. 19-23. IEEE Computer Society Press, Hongkong, China, November 14th-16th 2005.


Software The Dynamite software package for the checkpointing of PVM programs is available through http://www.science.uva.nl/research/scs/Software Development of "Roche" binary evolution web interface: http://carol.wins.uva.nl/~spz/act/roche/roche.html Development of "McScatter" hybrid software package for Monte-Carlo dynamics with binary evolution. http://manybody.org/McScatter.html MODESTA is a dedicated parallel computer platform for MOdeling DEnse STellar systems in Amsterdam. The machine became available for production on 2 October 2004 and is now working full time. Current performance is 1.01TFLOPs sustained speed. MODESTA has a dedicated web site: http://modesta.science.uva.nl/ Cooperation within ASCI The PhD thesis research by Iskra and many of our other research projects make heavy use of the DAS-2 supercomputer. K. Kruszynski; R. van Liere and J.A. Kaandorp: Quantifying performance of 3D skeletonization algorithms, in IASTED International Conference on Visualization, Imaging, and Image Processing (VIIP 2005), Benidorm, Spain, September 2005. (TUE-W&I-vis) L. Abrahamyan; J.A. Schaap; A.G. Hoekstra; D.P. Shamonin; F.M.A. Box; R.J. van der Geest; J.H.C. Reiber and P.M.A. Sloot: A Problem Solving Environment for Image-Based Computational Hemodynamics, in V.S. Sunderam; G.D. van Albada; P.M.A. Sloot and J.J. Dongarra, editors, Computational Science - ICCS 2005: 5th International Conference, Atlanta, GA, USA, Proceedings, Part I, in series Lecture Notes in Computer Science, vol. 3514, pp. 287-294. Springer, Berlin, Heidelberg, May 2005. ISBN 3-540-26032-3. (UL-LUMC-lkeb) In the BSIK funded VL-e project we collaborate with a number of ASCI groups affiliated with VU and UvA. In the NWO funded project DIME we collaborate with UL-LUMC-lked on a Problem Solving Environment for Computational Hemodynamics, and its potential use in pre-operative planning and training.


26

2.2 B: Large scale distributed Information Systems and Embedded Systems

2.2.1 Contribution of UvA-FdNWI-caps In the SNE group we work on different topics covered by Theme B2. In the following paragraphs we describe the research carried out in the context of both Advanced Networking and Grid middleware and workflow Advanced Networking Optical networks modeling and brokering In 2005 we focused on optical network modeling and resource brokering. In multi-domain optical networks the successful setup of an end-to-end connection depends on the existence of a coherent view of the available resources and their inter-connections. We researched suitable models for optical network description: we focused on extending the ITU G.805 standard and on semantic web techniques based on RDF. This latter model was used as basis for a prototype resource brokering system based on web services. We also looked at the use of link-local addressing with Zero configuration technology as a way to transparently setup the IP connectivity of end nodes in an optical environment. We participated to both iGrid2005 and SC2005 with demonstrations and proof of concepts for the link-local addressing and the resource brokering activities.

Network monitoring In 2005 we pursued two main activities in the field of network monitoring and testing : network monitoring in Grids and network monitoring on SURFnet6. In the Grid environment the traditional monitoring infrastructure and tools cannot be used as is, given they must rely on direct access to the end hosts. We developed a test infrastructure for the DAS-2 cluster that makes use of the Grid scheduling algorithms provided by MPICH-G2, the Globus variant of MPI. The periodic tests give information on the network health among the Grid clusters. The SURFnet6 network, the new generation SURFnet network, was installed and commissioned in 2005. We developed a suite of network performance tests to validate the performance of this new network, in particular at the 10GE level. Authorization Concepts and Architectures Within the realm of our research into the concepts, feasibility and efficiency of Generic AAA based mechanisms, the focus in the year 2005 was on the exploration of token-based sequences to manage and control network resources. The token sequence is one of 3 fundamental authorization sequences described in RFC2904. This sequence allows separation between acquiring and using an authorization, and therefore allows faster access enforcement. At SuperComputing ’05 we were successful in demonstrating 2 implementations of this sequence. In one implementation, a network resource received a token via a separate control channel. Within another implementation a token was embedded in the IP dataflow. With the first implementation, a token was used to manage resource access to a transatlantic optical network link between Amsterdam and Chicago using an Optical Cross Connect switch. The other implementation used special developed microcode within a Network Processor Unit based network switch, which generated and recognized a token inside an IP packet. This switch was used to provide both real time access control and path selection dependent on the presence of a valid token (Figure 1). Both developments lead to publications.

Figure 1: Token-based architecture for Resource management, path selection, and access control


The results of SuperComputing 2004 were published, where our Generic AAA toolkit, in combination with a vendor based Dynamic Operating Support System, implemented the concept of a Network Service Plane, capable of driving multiple autonomous network control planes. Such concept is explored within the GGF GHPN group. A prototype Role Based Access Control and Authorization mechanism based on the Generic AAA toolkit was finished within the Collaboratory.nl project. This success resulted in our involvement in the follow-up project, CNL-III, which will make the prototype ready for deployment. Activities made several contributions towards the EU project EGEE where we were involved in developing mechanisms enabling authentication and authentication in gLite (LCAS/LCMAPS). We performed architectural work and performed threat analyses in the area of Grid Security. The group also contributed new authorization concepts within the EU Nextgrid project.

Based on experiences gained in a several projects, a GAP analyses was performed to identify new requirements for the Generic AAA toolkit for Optical Light Path provisioning. Based on our recognition in the area of AAA, we were asked to be involved in 3 EU project proposals to cover aspects of authorization. One of these projects got selected for further negotiations. Standards body activities: Within the Global Grid Forum, we established the Firewall Issues Research Group and a group member acts as co-chair Grid Middleware en Workflow Management Grid-based Virtual Laboratory: - Distributed Services Discovery In the year 2005, we have started to investigate the Distributed Services Discovery of services for Grid-based Virtual Laboratory based on Peer-to-Peer Networks, we started by looking into the problem of classifying the services that have to be distributed and discovered by the targeted system. There are several ways to achieve the classification of the services using pattern recognition, Natural Languages, or Ontology based classifiers. We have decomposed this work into a number of steps: Survey the existing classifiers, model the targeted services, and evaluating the classifiers. Currently, we are investigating the classification problem in both structured and unstructured network. - Multi-layer Parameter sweep application Parameter sweep application has recently acquired more attention from the grid computing community since it is the potential candidate, which will require the non-trivial capabilities promised by the grid environment. As early experiment with parameter sweep applications has shown that, some of these promises from grid environment are indeed carried into effect, more demanding parameter sweep applications came into play such as parameter sweep workflows. This research task will focuses on analyzing strategies for scheduling layers of parameter sweep tasks within a workflow of parameter sweep into available resources on the grid. The approach will be to conduct simulation of selected algorithms for scheduling parameter sweep applications to set of available grid resources. The goal of this task is to evaluate existing strategy of scheduling parameter sweep workflow. Based on existing work on Critical-Path and Priority Based Algorithms for Scheduling. Assumption used in this paper is that layers of parameter sweep applications in a workflow do not communicate. Though it is indeed the case within one layer of parameter sweep task, it is no longer hold when we have layers of parameter sweep task in a workflow.

- Agent Technology We started also investigating how agent technology can be used in developing generic VL-e framework. Agent technology provides a suitable way to decompose and encapsulate complex control intelligence. We will study how agent technology facilitates the interoperability and integration among different workflow systems. External Projects VL-e 2004-2009 (Bsik project): WTCW, UvA, Amolf, Nikhef, CWI, SARA, VU, TUD, IBM, LigicaCMG, Philips, FEI GigaPort (Bsik project): UvA, Telematica Instituut , TNO StarePlane (NWO project, SURFnet and NORTEL): Collaboratory.nl : UvA, Telematica Instituut, Philips, FEI Company, DSM Lucifer: UvA, Surfnet, Sara


28

Papers in international Journals Curti, C., Ferrari, T., Gommans, L., van Oudenaarde, S., Ronchieri, E., Giacomini, F., Vistoli, C., On Advance Reservation of Heterogeneous Network Paths Future Generation Computer Systems Journal Vol. 21 Issue 4 (2005), pp. 525-538 S.M.C.M. van Oudenaarde, Z.W. Hendrikse, F. Dijkstra, L.H.M. Gommans, C.T.A.M. de Laat, R.J. Meijer An Open Grid Services Architecture Based Prototype for Managing End-to-End Fiber Optic Connections in a Multi-Domain NetworkHigh-Speed Networks and Services for Data-Intensive Grids: the DataTAG Project, special issue, Future Generation Computer Systems, volume 21 issue 4 (2005) Robert Grossman, Yunhong Gu, Xinwei Hong, Antony Antony, Johan Blom, Freek Dijkstra, and Cees de Laat, Teraflows over Gigabit WANs with UDTPublished in Future Generation Computer Systems, Vol. 21, Issue 4, pp.501-513, April 2005. Bas van Oudenaarde, Zeger Hendrikse, Freek Dijkstra, Leon Gommans, Cees de Laat, Rob Meijer, Dynamic paths in multi-domain optical networks for grids Published in Future Generation Computer Systems, Vol. 21, Issue 4, pp. 539-548, April 2005. Catalin Meirosu, Piotr Golonka, Andreas Hirstius, Stefan Stancu, Bob Dobinson, Erik Radius, Antony Antony, Freek Dijkstra, Johan Blom, Cees de Laat, Native 10 Gigabit Ethernet experiments over long distances Published in Future Generation Computer Systems, Vol. 21, Issue 4, pp. 457-468, April 2005. Freek Dijkstra, Jeroen van der Ham, Cees de Laat, Using Zero Configuration Technology for IP addressing in Optical Networks In Press (accepted by Future Generation Computer Systems, Feature topic iGrid 2005), First draft submitted October 2005, revision submitted February 2006. Paola Grosso, Pieter de Boer, Linda Winkler, The network infrastructure at iGrid2005: lambda networking in action In Press (accepted by Future Generation Computer Systems, Feature topic iGrid 2005), First draft submitted October 2005, revision submitted February 2006. Jeroen van der Ham, Freek Dijkstra, Franco Travostino, Bert Andree, Cees de Laat Using RDF to Describe Networks", In Press (submitted for Future Generation Computer Systems, Feature topic iGrid 2005), First draft submitted October 2005, revision submitted February 2006. Leon Gommans, Bas van Oudenaarde, Freek Dijkstra, Cees de Laat, Tal Lavian, Inder monga, Arie Taal, Franco Travostino, Alfred Wan Applications Drive Secure Lightpath Creation across Hetrogenous Domains In Press (accepted by IEEE Communications Magazine, Feature topic Optical Control Planes for Grid Networks: Opportunities, Challenges and the Vision), First draft submitted June 2005, revision submitted December 2005 Contributions to international conference proceedings Z. Zhiming, A.S.Z Belloum, Peter Sloot, and Bob Hertzberger Agent technology and scientific workflow management in an e-Science environment The 17th IEEE international conference on Tools with Artificial Intelligence, Hong Kong, China, Nov. 14 -16, 2005. Z. Zhiming, A.S.Z Belloum, Cees de Laat, and Bob Hertzberger Dynamic workflow in a Grid enabled Problem Solving Environment, the 5th International Conference on Computer and Information technology (CIT2005), Shanghai, China, Sep. 21-23, 2005. Z. Zhiming, A.S.Z Belloum, Adianto Wibisono, Frank Terpstra, Piter T. de Boer Peter Sloot, and Bob Hertzberger Scientific workflow management: between generality and applicability The 5th international conference on quality software, (to appear), Melbourne, Australia, Sep. 19 -20, 2005. I. Morozov, I. Shoshmina, A. Evlampiev, E.Stankova, A. Bogdanov, A. Luzan, D. Malashonok, I. Valuev, V. Korkhov. Experience in setting up an experimental Grid testbed for heavy scientific applications, All-Russian Scientific Conference on Scientific Service in Internet: distributed computing technologies, Abrau-Durso, Sep. 19-24, 2005 S. de Ridder, A.S.Z Belloum, and L.O. Hertzberger “Grid Service based Collaboration for VL-e: Requirements, Analysis & Design” European Grid Conference, February 14-16 2005, Science Park Amsterdam, The Netherlands

Morozov, I. Shoshmina, A. Evlampiev, E.Stankova, A. Bogdanov, A. Luzan, D. Malashonok, I. Valuev, V. Korkhov Experience in setting up an experimental Grid testbed for heavy scientific applications, All-Russian Scientific Conference on Scientific Service in Internet: distributed computing technologies, Abrau-Durso, Sep. 19-24, 2005.


Yuri Demchenko, Leon Gommans, Cees de Laat, Using VO concept for managing dynamic security associations, accepted paper for 21st IFIP International Information Security Conference Security and Privacy in Dynamic Environments May 22 - May 24, 2006 Karlstad University, Karlstad, Sweden Demchenko, Y., L. Gommans, C. de Laat, B.Oudenaarde, A. Tokmakoff, M. Snijders, Job-centric Security model for Open Collaborative Environment Proceedings 2005 International Simposium on Collaborative Technologies and Systems (CTS2005 ) May 15-19, 2005, Saint Louis, USA. - IEEE Computer Society, ISBN: 0-7695-2387-0. - Page 69-77. Yuri Demchenko, Leon Gommans, Cees de Laat, Bas Oudenaarde, Web Services and Grid Security Vulnerabilities and Threats Analysis and Model Accepted paper to Grid 2005 - 6th IEEE/ACM International Workshop on Grid Computing. - November 13-14, 2005. Leon Gommans, Cees de Laat, Robert Meijer, Token Based path authorization at Interconnection Points between Hybrid Networks and a Lambda Grid IEEE GRIDNETS2005 proceedings, ISBN 0-7803-9277-9. Yuri Demchenko, Leon Gommans, Cees de Laat, Bas Oudenaarde, Andrew Tokmakoff, Martin Snijders, Rene van Buuren, Security Architecture for Open Collaborative Environment Advances in Grid Computing EGC 2005: European Grid Conference, Amsterdam, The Netherlands, February 14-16, 2005, Revised Selected Papers, LNCS, Springer Verlag, Volume 3470, page 589, (2005) Jos Vrancken, Karst Koymans Intelligent Complexity in Internet Addressing SchemesProceedings of the IEEE International Conference on Systems, Man, and Cybernetics 2005 Experimental Software This program has been designed to minimize the effort needed by scientists to develop Grid-enabled application GVLAM: Data driven grid-enabled workflow management systems. 2000-2006: Java, C++, GT2.4, Web Services, Job Farming, Linux OS. This program has been designed to minimize to help the scientist to move file across geographically distributed storage resource using a simple Graphical user Interaface. Vbrowser: Data driven grid-enabled workflow management systems. 2005-2006: Java, FTP, SRB, HTML Cooperation within ASCI VL-e consortium include a number of ASCI members UvA, VU, TUD. We can consider the work developing the generic VL-e middleware as a collaboration of ASCI members.

2.2.2 Contribution of UL-LIACS Integration, Analysis and Logistics (DIAL) project Within the CMSB research groups from Leiden University, Leiden University Medical Center, TNO Prevention & Health, Vrije Universiteit Amsterdam, Vrije Universiteit Medical Center, and Erasmus Medical Center common diseases, such as cardiovascular disease, Alzheimer's, diabetes and rheumatism are studied. By combining the vast knowledge on a number of disease areas with the latest research technologies including genomics and bioinformatics, the aim is to elucidate the causes of various common diseases and apply these new insights in the development of new methods for diagnosis, new drugs and new means for prevention. During 2005 the main themes of DIAL were: · Further development of the functionality of the DIAL CGH Database. · Beta-testing of the database · Initiating of research usage of the database. External Projects Integration, Analysis and Logistics (DIAL) project 2003-2007, the Netherlands Genomics Initiative. Goals: To interface local databases; to establish a central database for experimental information; to develop biostatistics for the interlinking of Epidemiological and Systems Biology approaches; to implement commercial and develop in-house bioinformatics for the internal and external data mining and integration of all projects; to collaborate, within the BioASP framework, on the development of biostatistics, bioinformatics and GRID-based Virtual Laboratory Software; to develop education in biostatistics and bioinformatics.


30

Experimental software E.M. Bakker & F.J. Sicking, DIAL CGH Database (www.liacs.nl/~sicking/dial/). This database is a repository for storing raw MicroArray Data and quickly (re)calculating CGH Levels against the latest BAC and Oligo Sets. The data can be annotated in a MIAME compliant fashion. In the underlying Oracle database each user has a private database account, which means that data is private and not visible by others unless permitted. Performance analysis and design space exploration of embedded multi-processor Systems-on-chip and large-scale distributed embedded systems Given a number of relevant industrial media applications, and a media platform of type on-chip networked heterogeneous multiprocessor platform, model and program platform based architectures on a high level of abstraction, and with a high level of confidence in a reasonable time. External projects Project Artemisia 2004-2008, PROGRESS, 132 K + AIO + PostDoc, UVA CSA, TUD EWI-CE, TUD EWI –ST, TUE E-ESA The project includes the following: Firstly, translate given applications input-output equivalent process networks (an automated translator does exist for a restricted class of applications and is called Compaan. Secondly, model the platform, the architecture templates derived from it as well as the instantiation of these to architectures. Also on a high level of abstraction.Thirdly, provide low-level performance/cost numbers for the processing units on the architecture so that high-level performance/cost numbers are well calibrated and, hence, sufficiently accurate (a calibration tool/platform does exist for a restricted class processing units and is called Laura. A more advanced calibration tool/platform is to be developed/designed within Artemisia. Project MASSIVE 2000-2007, PROGRESS, 242 K + 4 AIO The aim of this project is to develop methods and tools to master the complexity of large scale embedded signal processing systems. The dataflow part of such a system (in particular distributed radio telescopes), has been modeled and analyzed with respect to performance in previous years. Currently the dataflow is modeled in an independent way, and the interfacing between dataflow and control flow is being modeled. Project Compaan and Laura 2001-2005, LIACS, 2 PHDs Automatic parallelizing of sequential programs and mapping on homogeneous and heterogeneous multi-processor embedded platforms. The parallelization is in terms of input-output equivalent Kahn Process Networks. Implementation is in SW, in HW, or in HW/SW. The related tools are Compaan, Laura, and Espam. Project Trader 2004-2008, ESI, 1 AIO, TUD EWI Modern systems such as household appliances, DVD players, PCs, medical X-ray imaging systems, printers, advanced vehicles, and airplanes rely increasingly on software, in particular for system integration. Embedded software monitors the whole system taking care that the system accomplishes more than its parts would. In such software intensive systems, reliability is of prime importance. Project Neva 2005-2009, EU, 50K + 2 AIO Is a European MEDEA+ collaboration between ACE, Bull Certess, LETI/CEA, Universiteit Leiden, Philips, Silicomp, STMicroelectronics, TIMA/INPG, and VERIMAG/UJF (Project 2A703: Networks on Chip Design Driven byVideo and Distribution Applications). Circuits for electronic devises are becoming so complex that they are expected to reach a billion transistors by the end of 2008. Traditional silicon chip architectures are nearing the limit of their performance in such applications, so the NEVA project was set up to introduce innovative network-on-chip designs, based on multiple processors and asynchronous circuitry. The goal is to allow designers and application engineers to cope with emerging applications resulting from multimedia/communications convergence. In first instance, datastream applications – mainly from video environments – will be used as drivers, with a targeted computing power of around one-giga operations per second per chip. Papers in international journals Turjan, A, Kienhuis, B., and Deprettere, E. “Solving Out-of-Order Communication inKahn Process Networks”, In. Journal of VLSI Signal Processing, Vol. 40, No. 1, pp. 7-18, May 2005 Contributions to international conference proceedings Nikolov, H., Stefanov, T., and Deprettere, E. “Modeling and FPGA implementation of Applications using parameterized Process Networks with Non-Static Parameters", In Proc. "13th IEEE Symposium on Field-Programmable Custom Computing Machines (FCCM'05)", pp. 255-263, Napa, California, USA, Apr. 18-20, 2005.


J. Lemaitre, Sylvain Alliot and Ed Deprettere Behavioral specification of control interface for signal processing applications IEEE Intl Conf. on Application-Specific Systems, Architectures, and Processors (ASAP<92>05)Samos, Greece, July 2005 Claudiu, Z., Kienhuis, B., and Deprettere, E., “Communication Synthesis in a multiprocessor environment”, in Proc. Field-Programmable Logic and Application Conference (FPL’05) Tampere, Finland, Aug. 24-26, 2005. Claudiu, Z., Kienhuis, A., and Deprettere, E. “Expression Synthesis in Process Network generated by Laura”, In Proc. 16th IEEE Int. Conf. on Application-specific, Systems, Architectures, and Processors (ASSAP’05), July 23-25, 2005. . Experimental software Nikolov, H., Stefanov, T., and Huang, K.2005, Java, Linux This is a tool suite for automatic platform synthesis and application mapping for multiprocessor Systems On-Chip. Application and PlatformAre specified on a level of abstraction above the RTL level. The tool suit transforms both to the RTL level in an automated way. Software and manual available.

2.2.3 Contribution of TUE-EE-dmes Embedded multi-media systems: architectures, models, programming and design The focus of our work is on mapping applications onto multiprocessor systems. System level design methods allow the systematic development of abstract executable models. These models are used to verify correctness and performance properties of the system. Based on the analysis results, system designers can take well-founded design decisions about the architecture of the system, hardware/software partitioning, the choice of processors, etc. Multiprocessor implementation platforms also require novel programming techniques, and techniques to map executable specifications onto multiprocessor systems. These techniques must optimize execution time, memory usage, and energy usage and allow trade-offs. The result is implementation-level code for the application optimized towards the intended multiprocessor platform. These programming techniques build upon compiler technology that covers the final step to the hardware. We set up an initial mapping trajectory for multi-media applications targeting multi-processor architectures. As a basic model of computation, we take Kahn process networks (KPN) and its derivates such as synchronous dataflow (SDF). We developed efficient techniques for analyzing the throughput of SDF graphs, and the entire Pareto space of throughput-buffer size trade-offs. We have identified the concept of application scenarios that capture common execution modes of a system implementation from the resource usage point of view. The mapping trajectory can use application scenarios to reduce energy consumption and to improve performance. We have also developed a hybrid VLIW/SIMD processor template, and design-space exploration techniques for VLIW and SIMD processors. For the longer term, embedded systems will evolve into ambient systems. Ambient systems are in essence nothing but networked embedded systems operating in a highly dynamic environment. More and more embedded systems have communication capabilities. This networking dimension affects the architectures, the design flow, the run-time systems, and the underlying computational models we are working on. We are looking at the integration of this dimension in our work. External projects Betsy: BEing on Time Saves energY Continuous multimedia experiences on networked hand-held devices, 2004-2007. EU/IST/FP6, M€ 4.4, Philips research, CSEM, IMEC, ISI, MDH/TU Kaiserslautern, Siemens C-lab, TUE, en de University of Cyprus The aim of the BETSY project is to have multimedia streams on wireless hand-held devices seamlessly adapted to fluctuating network conditions and available terminal resources while reducing the energy consumption of the stream processing. This way the user can enjoy true multimedia experiences with freedom of movement in a networked home or at

oper

atio

ns-1

qual

ity-1

qual

ity-1


32

any hot-spot. To achieve this, we need to be able to make trade-offs between the use and consumption of network and terminal resources, such as bandwidth use, CPU consumption, memory needed and power consumption by the terminal, while guaranteeing end-to-end timeliness - required for streaming data. PreMaDoNa Predictable Matching of Demands on Networked Architectures, 2004-2009, Progress/STW, M€ 1.97, Philips, CMG The PreMaDoNa project focuses on the design and implementation of NoC-based platforms for multi-media applications. The research challenges mainly originate from the increasing complexity of multi-media applications and the ever-shortening design time to realise multi-media systems. The major research objective is the following: Being able to design NoC-based real-time system in a predictable way such that non-functional properties can be guaranteed, while still being able to dynamically match quality with the available resources. PROgramming Multi-processor Embedded multi-media Systems (PROMES) 2002-2008, NWO, k€ 295 PROMES focuses on the development of both a sound theoretical framework and a programming environment for multi-media applications building on top of multi-processor systems. Important is the study of task-level analysis techniques that provide insight in concurrency-, timing-, and energy-related properties at the specification level without fully implementing an application. SmartCam 2002-2007. STW/progress, 1.4 M€, Philips Natlab, Philips CFT, TNO-FEL, In3D, HP Bristol labs, TUD-TNW-tn-ph The SmartCam project investigates low-cost one-chip Smart Camera solutions, contributing to a quantitatively guided design trajectory. In particular, we investigate the impact of current applications, and we try to define relevant architectural parameters and to develop an architectural template. Other aims are to enhance and integrate existing application mapping environments for SIMD and ILP processors. Fexible Application Mapping Environment (FAME) 2003-2008, NWO, 250 k€, Leiden University, Delft University of Technology, TUD-ITS-me-ce FAME aims to obtain low power solutions when mapping applications on processor platforms. We like to achieve this goal by creating a compiler infrastructure capable of performing source-code transformations. We propose a dynamic approach in combination with analytic pruning of the transformation search space in order to find the best low-power optimizations. Beyond the Ordinary: Design of Embedded Real-time Control (BODERC) 2003-2008. Senter, 2.5 M€, Océ Technologies, Philips CFT, AAS, Imtech ICT, Chess iT, Katholieke Universiteit Nijmegen, Universiteit Twente. The Boderc project focuses on distributed embedded real-time controllers of complex systems. An Océ printer is taken as a case-study and acts as a driver for the project. The target is an integral approach for a systematic architectural design, modeling, analysis, and validation methodology for such heterogeneous systems. Contributions to Books Bekooij, M. van; Meerbergen, J.L. van; Hoes, R.J.H.; Moreira, O.; Poplavko, P.; Pastrnak, M; Mesman, B.; Mol, J.D.; Stuijk, S.; Gheorghita, V.S.: Dataflow Analysis for Real-Time Embedded Multiprocessor System Design. In Dynamic and robust Streaming in and between Connected Consumer-Electronic Devices, Chapter 4. ISBN 1-4020-3453-9, ed. P. van der Stok; Springer, Berlin, Germany, 2005, pp. 81-108. Huang, J.; Voeten, J.P.M.; Florescu, O.; Putten, P.H.A. van der; Corporaal, H.: Chapter 8, Predictability in real-time system development. Advances in Design and Specification Languages for SoCs, Chapter 8. ISBN 0-387-26149-4, ed. P. Boulet; Kluwer Academic Publishers, 2005. Papers in international journals Aa, T. van der; Jayapala, M.; Barat, F.; Deconinck, G.; Lauwereins, R.; Corporaal, H.; Catthoor, F.: Instruction buffering exploration for low energy embedded processors. JEC 1, nr. 3, 2005, pp. 341-351. Ciordas, C.; Basten, T.; Radulescu, A.; Goossens, K.G.W.; Meerbergen, J.L. van: An Event-based Monitoring Service for Networks on Chip. ACM Transactions on Design Automation of Electronic Systems 10, nr. 4, 2005, pp. 702-723. Gheorghita, V.S.; Corporaal, H.; Basten, A.A.: Iterative Compilation for Energy Reduction. JEC 1, nr. 4, 2005. Jayapala, M.; Barat, F.; Aa, T. van der; Catthoor, F.; Corporaal, H.; Deconinck, G.: Clustered Loop Buffer Organization for Low Energy VLIW Embedded Processors. IEEE Transactions on Computers 54, nr. 6, 2005, pp. 672-683.


Contributions to international conference proceedings Aa, T. van der; Jayapala, M.; Barat, F.; Corporaal, H.; Catthoor, F.; Deconinck, G.: A High Level Memory Energy Estimator based on Reuse Distance. Proceedings of 3rd workshop on Optimizations for DSP and Embedded Systems, ODES 2005 together with CGO 2005., 20-23 March 2005. Aa, T. van der; Corporaal, H.; Catthoor, F.; Deconinck, G.: Combining data and instruction memory energy optimizations for embedded applications. Proceedings of the 3rd workshop on Embedded Systems for Real-Time Multimedia 2005, 22-23 September 2005, ISBN 0-7083-9347-3, ed. M. Miranda; S. Ha; IEEE, 2005, pp. 121-126. Avasare, P.; Nollet, V.; Mignolet, J.-Y.; Verkest, D.; Corporaal, H.: Centralized End-to-End Flow Control in a Best-Effort Network-on-Chip. Proceedings of the EMSOFT conference 2005, ISBN 1-59593-091-4, IEEE, 2005, pp. 17-20. Barat, F.; Aa, T. van der; Jayapala, M.; Deconinck, G.; Lauwereins, R.; Corporaal, H.: Methodology for building processor design space exploration frameworks. Proceedings of 3rd Workshop on Optimizations for DSP and Embedded Systems (ODES 2005), 20-20 March 2005. Caarls, W.; Jonker, P.P.; Corporaal, H.: Skeletons and Asynchronous RPC for Embedded Data- and task Parallel Image Processing. Proceedings of the 9th IAPR Conference on Machine Vision Applications 2005, 16-18 May 2005, ed. K Ikeuchi, 2005. Fatemi, S.H.; Corporaal, H.; Basten, A.A.; Kleihorst, R.; Jonker, P.: Designing Area and Performance Constrained SIMD/VLIW Immage Processing Architectures. Proceedings of the 7th Int. Conference of Advanced Concepts for Intelligent Vision Systems. ACIVS 2005, 20-23 September 2005, ISBN 3-540-29032-X, ed. J. Blanc-Talon; W. Philips; D. Popescu; P. Scheunders; Springer, Berlin, 2005, pp. 689-696. Florescu, O.; Voeten, J.P.M.; Corporaal, H.: Property-Preserving Synthesis for Unified Control- and Data-Oriented Models. Proceedings of the Forum on Specification & Design Languages (FDL) 2005, 27-30 September 2005. Geilen, M.C.W.; Basten, T.; Theelen, B.D.; Otten, R.H.J.M.: An Algebra of Pareto Points. Proceedings of the 5th Application of Concurrency to System Design 2005 conference, 6-9 June 2005, ISBN 0-7695-2363-3, ed. J. Desel; Y. Watanabe; IEEE Computer Society Press, Los Alamitos, 2005, pp. 88-97. Geilen, M.C.W.; Basten, T.; Stuijk, S.: Minimising Buffer Requirements of Synchronous Dataflow Graphs with Model Checking. Proceedings of the 42nd Design Automation Conference, DAC 2005, 13-17 June 2005, ISBN 1-59593-133-3; ACM Press, New York, 2005, pp. 819-824. Gheorghita, V.S.; Stuijk, S.; Basten, T.; Corporaal, H.: Automatic Scenario Detection for Improved WCET Estimation. Proceedings of the 42nd Design Automation Conference, DAC 2005, 13-17 June 2005, ISBN 1-59593-133-3; ACM Press, New York, 2005, pp. 101-104. Gheorghita, V.S.; Grigore, R.: Constructing Checkers from PSL Properties. Proceedings of the 15h Int. Conference on Control Systems and Computer Science. 2, ISBN 973-8449-89-8, 2005, pp. 757-763. Gheorghita, V.S.; Basten, T.; Corporaal, H.: Intra-task Scenario-aware Volgage Scheduling. Proceedings of CASES'2005, 24-27 September 2005, ISBN 1-59593-149-X; ACM Press, New York, 2005, pp. 177-184. Heikkinen, J.; Cilio, A.; Takala, J.; Corporaal, H.: Dictionary-Based Program Compression on transport Triggered Architectures. Proceedings IEEE Int. Symposium on Circuits and Systems, 23-26 May 2005, ISBN 0-7803-8835-6, 2005, pp. 1122-1125. Heikkinen, J.; Takala, J.; Corporaal, H.: Dictionary-Based Program Compression on TTAs: Effects on Area and Power Consumption. Proceedings of IEEE Workshop on Signal Processing Systems., 2-4 November 2005, ISBN 0-7803-9334-1, 2005, pp. 479-484. Lambrechts, A.; Corporaal, H.; Robert, F.; Carrabina, J.; Raghavan, P.; Leroy, A.; Talavera, G.; Aa, T. van der; Jayapala, M.; Catthoor, F.; Verkest, D.; Deconinck, G.: Power breakdown analysis for a Heterogeneous NoC Platform running a video Application. Proceedings of IEEE 16th International Conference on Application-specific Systems, Architectures and Processors ASAP 2005, IEEE 2005. Lu, R.X.; Silva, C.; Ang, M.H.; Poo, J.A.N.; Corporaal, H.: A New Approach for Mechatonic System Design: Mechatronic Design Quotient (MDQ). Proceedings of the 2005 IEEE/ASME International Conference on Advanced Intelligent Mechatronics. 2005, pp. 911-915.


34

Marescaux, T.; Rangevall, A.; Nollet, V.; Bartic, A.; Corporaal, H.: Distributed congestion control for packet switched networks on chip. Proceedings of Parallel Computing Conference PARCO'05, 25-25 January 2005. Marescaux, T.; Bricke, B.; Debacker, P.; Nollet, V.; Corporaal, H.: Dynamic Time-Slot Allocation for QoS Enabled Networks on Chip. Proceedings of the 3rd Workshop on Embedded Systems for Real-Time Multimedia 2005; IEEE, 2005, pp. 47-52. Ovadia, I.; Ha, Y.; Corporaal, H.: Using Multiple Paths in NoCs for Guaranteed Resource Allocation and Improved Best Effort Performance. Proceedings of ProRISC 2005, 2005. Palkovic, M.; Corporaal, H.; Catthoor, F.: Global memory optimisation for embedded systems allowed by code duplication. Proceedings of International Workshop on Software and Compilers for Embedded Systems, SCOPES 2005, 2005, pp. 72-79. Palkovic, M.; Corporaal, H.; Catthoor, F.: Scenario creation for low power embedded systems. Proceedings of Architectures and Compilers for Embedded Systems ACES 2005, 2005, pp. 26-29. Palkovic, M.; Brockmeyer, E.; Vanbroekhoven, P.; Corporaal, H.; Catthoor, F.: Systematic preporocessing of data dependent constructs for embedded systems. Proceedings of 15th international workshop Integrated Circuit and System Design. Power and Timing Modeling Optimization and Simulation, PATMOS 2005, 2005, pp. 88-99. Seneclauze, M.; Blanch, C.; Bormans, J.; Geilen, M.C.W.; Basten, T.; Theelen, B.D.; Koulamas, C.; Papadopoulos, G.; Prayati, A.; Fohler, G.; Isovic, D.; Decotignie, J.D.; Papadopoulos, G.A.; Cheng, P. et.al.: The BETSY Project on Timeliness and Energy Aspects of Wireless Video Streaming. Proceedings of the International Workshop on Wireless Ad-hoc Networks. Electronic proceedings, 2005. Stuijk, S.; Basten, T.; Mesman, B.; Geilen, M.C.W.: Predictable embedding of large data structures in multiprocessor networks-on-chip. Proceedings of he 8th EUROMICRO Conference on Digital System Design, 30 August - 3 September 2005, ISBN 0-7695-2433-8, ed. C. Wolinski; IEEE Compuer Society Press, Los Alamitos, 2005, pp. 388-395. Terechko, A.; Garg, M.; Corporaal, H.: Evaluation of Speed and Area of Clustered VLIW Processors. Proceedings of the 18th Int. Conference on VLSI Design and the 4th Int. Conference on Embedded Systems Design, 3-7 January 2005, ISBN 0-7695-2264-5; IEEE Computer Society, 2005, pp. 557-563. Voeten, J.P.M.; Huang, J.; Florescu, O.; Theelen, B.D.; Corporaal, H.: Towards predictability in real-time embedded system design. Invited presentation. Proceedings of the Lorentz-ARTIST (Network of Excellence on Embedded Systems Design) Workshop Embedded Systems. 22-22 November 2005. Ykman-Couvreur, Ch.; Brockmeyer, E.; Nollet, V.; Marescaux, T.; Catthoor, F.; Corporaal, H.: Design-Time Application Exploration for MP-SoC Customized Run-Time Management. Proceedings of Systems on Chip Conference (SoC '05), ISBN 0-7803-9294-9, 2005, pp. 66-69. Abstracts Poplavko, P.; Basten, T.; Pastrnak, M; Meerbergen, J.L. van; Bekooij, M. van; With, P.H.N. de: Extended Abstract: Estimation of Execution Times of On-chip Multiprocessors Stream-oriented Applications. Formal Methods and Models for Codesign, 3rd ACM & IEEE Int. Conference, MEMOCODE 2005, proceedings., 11-14 July 2005, ISBN 0-7803-9227-2; IEEE Computer Society Press, Los Alamitos, 2005, pp. 251-252. Stuijk, S.; Basten, T.; Mesman, B.; Geilen, M.C.W.: Predictable embedding of large data structures in multiprocessor networks-on-chip (extended ebstract). Proceedings of DATE 2005, 7-11 March 2005, ISBN 0-7695-2288-2, ed. N. Wehn; Benini; IEEE Computer Society Press, Los Alamitos, 2005, pp. 254-255. Experimental software Fatemi, H; Caarls, W; SmartCam Image Processing Library 2003 - 2005, C/Unix, C++/MPI/Linux, 1DC/IMAP, XTC/Xetal, StreamC/kernelC/Imagine. The library is based on algorithmic skeletons for image processing operations (low, intermediate and high level). By using this library, the programmer of an image processing application can easily parallelize the application and s/he does not have to handle the problems related to communication and synchronization. Stuijk, S; Multiprocessor design flow 2002 - now , C++/JAVA/Linux/Windows. Next-generation embedded multi-media systems will often be built on multi-processor systems to obtain high compute power at relatively low energy cost. Work is ongoing to implement a design flow to map streaming applications specified as


synchronous dataflow (SDF) graphs to a heterogeneous multi-processor systems. We aim at predictability with respect to timing behavior, while minimizing energy consumption. An important component of this design-flow is a tool to analyze SDF graphs, called SDF3. The tool also allows the generation of random test graphs, and it supports the visualization of graphs. The tool is available via http://www.es.ele.tue.nl/sdf3. Products of cooperation within ASCI Caarls, W.; Jonker, P.P.; Corporaal, H.: Skeletons and Asynchronous RPC for Embedded Data- and task Parallel Image Processing. Proceedings of the 9th IAPR Conference on Machine Vision Applications 2005, 16-18 May 2005, ed. K Ikeuchi, 2005. Fatemi, S.H.; Corporaal, H.; Basten, A.A.; Kleihorst, R.; Jonker, P.: Designing Area and Performance Constrained SIMD/VLIW Immage Processing Architectures. Proceedings of the 7th Int. Conference of Advanced Concepts for Intelligent Vision Systems. ACIVS 2005, 20-23 September 2005, ISBN 3-540-29032-X, ed. J. Blanc-Talon; W. Philips; D. Popescu; P. Scheunders; Springer, Berlin, 2005, pp. 689-696. Fatemi, H; Caarls, W; SmartCam Image Processing Library 2003 - 2005, C/Unix, C++/MPI/Linux, 1DC/IMAP, XTC/Xetal, StreamC/kernelC/Imagine.The library is based on algorithmic skeletons for image processing operations (low, intermediate and high level). By using this library, the programmer of an image processing application can easily parallelize the application and s/he does not have to handle the problems related to communication and synchronization.

2.2.4 Contribution of UT- EEMCS-dacs Management and measurement of operational networks In this activity, we work on methods and techniques to improve network performance and dependability, based on measurements taken in a life network. Today, the network we focus on is primarily the fixed internet. Using active and passive measurement techniques, critical network parameters are derived (using statistical techniques). Simple models are fed with these parameters; model evaluations help in decision making in order to imrove network performance and dependability. Decisions are put into effect using network management techniques (e.g. SNMP or web-services based). Intrusion detection (anomaluy based) is a new activity we are starting in this field. Mobile and wireless communications This activity focuses on the development of architectures, protocols, and algorithms for future wireless networks. Aspects that play a role here are for instance access mechanisms, routing, service discovery, and all kinds of traffic control aspects, such as scheduling, congestion control, power control, etc. New mechanisms are designed, and the design in analyzed using analytical models, simulation models, and prototype implementations. The current work concentrates on self-organizing ad hoc networks, that optimize their operation based on information regarding other nodes in the network and the environment. Design and analysis of networked embedded systems Networked embedded systems are starting to appear as real systems. Networked embedded systems typicallly have to operate under severe resource constraints; furthermore, best effort services are typically not appropriate. Hence, new lightweight protocols have to be designed and validated for their correctness and performance and dependability properties. Our research in this area primarily focusses on specifications of such systems, followed by numerical evaluation of critical performance and dependability parameters. Our specification and evaluation techniques use classical techniques known from queueing theory, but also more advanced techniques pairing stochastic analysis with model checking. We started work on using the DAS machine for distributed evaluation of these very large models. External projects MC MC, 2004-2008, NWO, 350 kEU, -.This project focusses on specification and numerical evaluation of infinite-state systems, using stochastic model checking technique. Techniques known for finite-state systems model checking, will be adapted and enhanced for infinite-state systems. This is a cooperation project between the FMT group (Brinksma) and DACS, both at the UT. VOSS2 2005-2007, NWO, 55 kEU, -.This project facilitates the cooperation between two universities in the Netherlands (UT (Haverkort, Katoen), RUN (Vaandrager)) and 3 universities in Germany (Bonn (Baier), Saarbruecken (Hermanns) and Muenchen (Siegle)).


36

MoQS 2006-2008, NWO, 500 kEU, -. Cooperation project (NWO/FOCUS) with the FMT group (Stoelinga) at UT, in which work on component-based design and evaluation of large systems is being performed. Tool support for that purpose is being developed as well. VeriGRID 2006-2008, NWO, 500 kEU, -.Cooperation project (NWO/FOCUS) between TU/e (Groote), CWI (Van der Pol) and UT, in which large-scale distributed algorithms are developed for model checking purposes, in order to increase the size of the systems that can be practically analysed with these techniques by at least one order of magnitude. Use of the ASCI DAS-3 is foreseen. QoSPN@Home 2004-2008, EZ/Senter, 400 kEuro (DACS budget), -. Cooperation project between TUDelft (Niemegeers/Van Mieghem) and UT, on the design and analysis of QoS-aware personal networks. Such networks are overlay networks consisting of ad hoc nodes (such as in a personal area network) and potentially fixed nodes. Freeband/Awareness BSIK, 500 kEuro (DACS). Cooperation project between UT and Lucent Tehnologies, Telematica Instituut, Ericsson, Roessingh R&D, Yucat, TMS, and WMC, researching and developing context-aware services and networks. The DACS group investigates context discovery for ad hoc networks in this project. Freeband/PNP2008 SIK, 500 kEuro (DACS) . Cooperation project between UT and TUDelft, Philips, KPN, WMC,and TNO, to research and develop a Personal Network, focussing on the prototyping of a Personal Mobile Gateway. Gigaport NG Research on Networks, 2005-2007, BSIK, 150KE, UvA & TUD. Within this project the UT investigates self-management of hybrid IP/ Optical networks, and the measurement of traffic flows within such networks. The project is managed by SURFnet, and supports the introduction of SURFnet6. Emanics 2006-2009, EU, 310 KE, -.European Network of Excellence (NoE) for the Management of Internet Technologies and Complex Services. This NoE addresses the scalability, dynamics, security and automation challenges that emerge towards the management plane of the future Internet and complex services running on top of it. UT coordinates the research within this NoE. Contributions to Books Parhonyi, R., Nieuwenhuis, L.J.M., Pras,A. The rise and fall of micropayment systems. In T. Lammer (Ed.), Handbuch E-Money,E-Payment & M-Payment (pp. 343-361). Heidelberg: Springer Verlag (ISBN 3-7908-1651-5). Papers in international journals Ait Yaiz, R., Heijenk,G.J. Providing QoS in Bluetooth. Cluster computing, (ISSN 1386-7857), 8, 223-231. Baier, C., Hermanns, H., Katoen, J.P., Haverkort, B.R.H.M. Efficient computation of time-bounded reachability probabilities in uniform continuous-time Markov decision processes. Theoretical computer science, (ISSN0304-3975), 345, 2-26. Baier, C. , Haverkort, B.R.H.M., Hermanns, H., Katoen, J.P. Model Checking meets Performance Evaluation. Performance evaluation review,(ISSN 0163-5999), 32, 10-15. Bell, A., Haverkort,B.R.H.M. Sequential and distributed model checking of Petri nets. International journal on software tools for technology transfer, (ISSN 1433-2779), 7(1), 43-60. Boer, P.T. de, Kroese, D.P., Mannor, S., Rubinstein, R. A Tutorial on the Cross-Entropy Method. Annals of operations research, (ISSN 0254-5330), 134, 19-67. Boer, P.T. de. Rare-event simulation of non-Markovian queueing networks using a state-dependent change of measure determined using cross-entropy. Annals of operations research, (ISSN 0254-5330), 134, 69-100. Clemm, A., Festor, O., Pras, A. A Report on IM 2005. Journal of network and systems management, (ISSN 1064-7570), 13(3), 351-354.


Haverkort, B.R.H.M., Katoen, J.P. Performance and Verification. Performance evaluation review, (ISSN 0163-5999), 32(4),3. Karagiannis, G., Fu, X., Schulzrinne, H., Bader, A., Hogrefe, D., Kappler, C., Tschofenig, H., & Bosch, S. van den. NSIS: A New Extensible IP Signaling Protocol Suite. IEEE communications magazine, (ISSN 0163-6804), 43(10), 133-141. Niemegeers, I.G.M.M., Heemstra de Groot, S.M. FEDNETS: Context-aware Ad-hoc Network Federations. Wireless personal communications, (ISSN 0929-6212), 33, 319-325. Pavlou, G., Pras, A. Introducing the Series on Network and Service Management. IEEE communications magazine, (ISSN 0163-6804), 43(10), 60-60. Contributions to international conference proceedings Bader, A., Karagiannis, G., Westberg, L., Kappler, C., Phelan, T., Tschofenig, H., & Heijenk, G.J. QoS Signaling Across Heterogeneous Wired/Wireless Networks: Resource Management in Diffserv Using the NSIS Protocol Suite. In S. Shen (Ed.), Proceedings of QShine 2005 .Orlando, Florida USA: IEEE Communications Society (ISBN 0-7695-2423-0). Bomhoff, M.J., Eyckelhof, C.J., Meent, R. van de , & Pras, A. Quarantine Net: design and application. In Application session proceedings of the 9th IFIP/IEEE Int. Symposium on Integrated Network Management . Nice: IEEE (ISBN 0-7803-9088-1). Brandhorst, C.J., & Pras, A. DNS: a statistical analysis of name server traffic at local network-to-Internet Connections. In C. Delgado Kloos & et al. (Eds.), Proceedings of the 11th Open European Summer School and IFIP WG6.4/6.6/6.9 Workshop (pp. 72-78) (ISBN 84-89315-43-4). Cloth, L. , & Haverkort, B.R.H.M. Hyperbolic PDE's for CSRL Model Checking: A Deja Vu. In Proceedings of the 7th Int.'l Workshop on Performability Modeling Computer and Communication Systems (pp. 19-22). Torino, Italy. Cloth, L. , Katoen, J.P., & Khattri, M. Model Checking Markov Reward Models with Impulse Rewards. In A. Dr. Bondavalli, B.R.H.M. Haverkort, & D. Prof. Tang (Eds.), Proceedings of the Int.'l Conference on Dependable Systems and Networks (pp. 722-731). Yokohama, Japan: IEEE Computer Society (ISBN 0-7695-2282-3). Cloth, L. , & Haverkort, B.R.H.M. Surviving Survivability Specifications. In A. Dr. Bondavalli, B.R.H.M. Haverkort, & D. Prof. Tang (Eds.), Supplemental Volume of the 2005 Int.'l Conference on Dependable Systems and Networks (pp. 70-71). Yokohama, Japan:IEEE Computer Society. Heijenk, G.J. , & Liu, F. (2005). Interference-based Routing in Multi-hop Wireless Infrastructures. In T. Braun, G. Carle, Y. Koucheryavy, & V. Tsaoussidis (Eds.), Proceedings WWIC 2005, 3rd Int.'l Conference on Wired/Wireless Internet Communications Vol. 3510. Lecture notes in computer science, (pp. 117-127). Xanthi, Greece: Springer Verlag (ISBN 3-540-25899-X). Jacobsson, M., Hoebeke, J., Heemstra de Groot, S.M., Lo, A., Moerman, I., Niemegeers, I.G.M.M., Munoz, L., Alutoin, M., Louati, W., & Zeghlache, D. A Network Architecture for Personal Networks. In P. Herhold & R. Schlager (Eds.), Proceedings of the 14th IST Mobile and Wireless Communication Summit . Dresden, Germany. Jehangir, A. , Heemstra de Groot, S.M. A Secure and Lightweight Ad-hoc Routing Algorithm for Personal Networks. In L. Heinzl & N.R. Prasad (Eds.), Proceedings of the Int.'l Conference on Wireless Personal Multimedia Communications (pp. 1998-2002). Aalborg, Denmark (ISBN87-90834-79-8). Karagiannis, G. , Hancock, R., Loughney, J., & Bosch, S. van den. Next Steps in Signaling: Framework. In Proceedings IETF'63 . Paris, France:IETF. Karagiannis, G. , Manner, J., Bosch, S. van den, & McDonald, A. NSLP for Quality-of-Service signalling. In Proceedings IETF'63 . Paris, France: IETF. Karagiannis, G. , Bader, A., Westberg, L., Kappler, C., & Phelan, T.. RMD-QOSM - The Resource Management in Diffserv QoS model. In Proceedings IETF'63 . Paris, France: IETF. Karagiannis, G. , Bader, A., & Westberg, L. Using IP as Transport Technology in Third Generationand Beyond Radio Access Networks. In H. de Meer & N. Bhatti (Eds.), Proceedings of 13th IEEE Int.'l Workshop on Quality of Service (IWQoS2005) Vol. 3552. Lecture notes in computer science, (pp. 372-375). Passau, Germany: Springer Verlag (ISBN 3-540-26294-6). Lo, A., Heijenk, G.J. , Niemegeers, I.G.M.M.. Evaluation of MPEG-4 Video Streaming over UMTS/WCDMA Dedicated Channels. In Proceedings of the 1st Int.'l Conference on Wireless Internet . Budapest, Hungary.


38

Lo, A., Heijenk, G.J. , Niemegeers, I.G.M.M. Performance Evaluation of MPEG-4 Video Streaming over UMTS Networks using an Integrated Tool Environment. In M.S. Obaidat, L. Marchese, J.L. Marzo, & F. Davoli (Eds.), Proceedings of the Symposium on Performance Evaluation of Computer and Telecommunication Systems . Philadelphia, USA: The Society for Modeling and Simulation Int.'l (SCS) (ISBN 1-56555-300-4). Parhonyi, R., Nieuwenhuis, L.J.M. , Pras, A. . Second generation micropayment systems: lessons learned. In M. Funabashi & A. Grzech (Eds.), Proceedings: Challenges of expanding Internet: E-commerce, E-business and E-government (pp. 345-359). Springer (ISBN 0-387-28753-1). Prasad, V., Jacobsson, M., Lo, A., Niemegeers, I.G.M.M., Heemstra de Groot, S.M. Architectures for Intra-Personal Network Communication. In M.M. Buddhikot & Lo R. Cigno (Eds.), Proceedings of the Third ACM Int'.l Workshop on Wireless Mobile Applications and Services on WLAN Hotspot . Cologne, Germany: Springer Verlag. Remke, A.K.I., Haverkort, B.R.H.M. Beyond Model-Checking CSL for QBDs: Resets, Batches and Rewards. In Proceedings of the 7th Int.'l Workshop on Performability Modeling Computer and Communication Systems (pp. 23-26). Torino, Italy: University of Torino. Remke, A.K.I. , Haverkort, B.R.H.M. , Cloth, L. Model Checking Infinite-State Markov Chains. In N. Halbwachs & L.D. Zuck (Eds.), Tools and Algorithms for the Construction and Analysis of Systems Vol. 3440. Lecture notes in computer science, (pp. 237-252). Edinburgh, UK: Springer Verlag (ISBN 10-3-540-25333-). Vulic, N, Heemstra de Groot, S.M., Niemegeers, I.G.M.M. 802.11 Modifications for WLAN-UMTS Integration at Radio Access. In L. Heinzl & N.R. Prasad (Eds.), Proceedings of the Int.'l Conference on Wireless Personal Multimedia Communications (pp. 1206-1210). Aalborg, Denmark (ISBN87-90834-79-8). Vulic, N, Heemstra de Groot, S.M., & Niemegeers, I.G.M.M. A Comparison of Interworking Architectures for WLAN Integration at UMTS Radio Access Level. In Proceedings of the First Int.'l Workshop on Convergence of Heterogeneous Wireless Networks, Budapest, Hungary. Vulic, N, Heemstra de Groot, S.M., & Niemegeers, I.G.M.M. Common radio resource management for WLAN-UMTS integration Radio Access Level. In P. Herhold & R. Schlager (Eds.), Proceedings of the 14th IST Mobile and Wireless Communication Summit, Dresden, Germany. Vulic, N, Heemstra de Groot, S.M., & Niemegeers, I.G.M.M.. UMTS-based Architectures for UMTS-WLAN Integration. In van W.C. Etten & R. Srinivassan (Eds.), Proceedings of the IEEE 12th Symposium on Communications and Vehicular Technology in the Benelux SCVT 2005 . Enschede: IEEE. Wanrooij, W. van, & Pras, A. Data on Retention. In J. Schoenwaelder & J. Serrat (Eds.), Proceedings of the 16th IFIP/IEEE Int.'l Workshop on Distributed Systems: Operations and Management (DSOM 2005) Vol. 3775. Lecture notes in computer science, (pp. 60-71). Barcelona, Spain: Springer Verlag (ISBN 3-540-29388-4). Wanrooij, W. van, & Pras, A. DNS Zones Revisited. In C. Delgado Kloos & et al. (Eds.), Proceedings of the 11th Open European Summer School and IFIP WG6.4/6.6/6.6/6.9 (pp. 84-92). Colmenarejo, Spain (ISBN 84-89315-43-4). Highlight


As highlight from the publication list, we would like to put forward the paper: Alexander Bell, Boudewijn R.H.M. Haverkort, Sequential and distributed model checking of Petri nets. International Journal on Software Tools for Technology Transfer, 7(1), 43-60 (Springer-Verlag). This paper, in essence, reports about a PhD project that was completed while I was still working at the RWTH; in fact, the PhD project finished in november 2003, the paper ws finalized in the beginning of 2004, and published in 2005. The paper describes an application /from computer science/ for large-scale parallel/distributed computing. What we see most often in computational science, is that clusters and grids are being used to solve problems in the physical sciences, like geology, astrology, or the engineering sciences, like aerodynamics, mechanics, and so on. Here, instead, we use a cluster of 26 dual processor nodes to attack large-scale model checking problems. Model checking is an exhaustive search and proof technique to verify the validity of formally specified properties in the context of a formally described system description. It is currently seen as one of the most promising automated correctness proof systems and is, for instance, widely employed in the context of embedded software and communication protocol design. However, "production model checking" does not emply parallel and distributedalgorithms so far. In our work, we developed parallel/distributed algorithms to verify correctness properties specified in a temporal logic called CTL (computational tree logic). The system is described using Petri nets (but other description formalisms could be use equally well). The real problem in disitributed model checking is that the state space of the system being described as Petri net is a priori unknown, and in any case, unstructured. Furthermore, the state space may easily contain many millions of states, so that it cannot be stored in main memory; this counts even more for the state-transition relation, which is, typically, 10 to 20 times larger than the state space itself. The unstructuredness of the state space makes the work division strategy very difficult. We developed a probabilistic work division concept based on hashing functions. We dealt with the enormous memory requirements by using disk-based techniques, as well as on-demand recomputation techniques; given a state, its possible predecessors are re-computed whenever necessary, instead of storing them. The algorithms have been implemented on the PARSECS cluster, a machine we built ourselves in 2000, for slightly less than DM 200.000,- (less than 100.000,- Euro). Cooling and appropriate power supply accounted for another DM 10.000,-. This Linux cluster comprises 26 dual Pentium III nodes (500 MHz), equipped with 512 MB main memory and a 40 GB local disk per node, connected via a switched 100 Mbps Erhernet (see figure); early in 2006, the cluster was put out of operation. The algorithms we developed scale surprisingly well: we achieved efficiencies of around 90% for small processor numbers, around 80% for26 processors, going down to some 70% for 52 prcoessors (probably due to intra-node bus contention). The larger the problems we addressed, the better the speed-ups and efficiencies. Using the same cluster, and similar algorithms, we recently also developed numerical solution procedures for Markov chains based on Petri net models. We were able to solve Markov chains with close to a billion states, meaning that we solve linear systems with up to a million unknowns, in a distributed, disk-based fashion. These are the largest Markov chains known publicly tohave been solved; Google might solve larger Markov chains in the context of their pagerank computations, but they did not (yet) report about that in the open literature.

2.2.5 Contribution of VU – WI Research in the group is concentrated in three areas, Large-Scale distributed Systems, Security, as well as Cluster and Grid computing. Large-Scale Distributed Systems We are concentrating on three strongly interrelated subareas. First, we are putting effort into developing a fully decentralized user-centric content delivery networks, called Globule. Globule handles automatic and adaptive replication of Web pages for the sake of performance and availability. A first version of Globule was released in December 2003. Various Web sites are currently hosted by Globule software. The current version is 1.3.2 and supports Windows as well as UNIX platforms. In 2005, we have been concentrating on research aimed at service-oriented architectures, replication of dynamically generated documents, and automated resource provisioning. Second, we are increasing our efforts concerning large-scale epidemic-based systems, covering wired as well as wireless networks. In particular, attention has been paid to protocols for decentralized membership management for very large networks, allowing nodes to regularly join and leave the system. Also, we have been looking at using epidemics for efficiently monitoring systems. These efforts are now leading to solutions for large-scale sensor networks, where we concentrate on debugging, monitoring, and security with emphasis on local-only solutions. Security This work concerns reliability and security in various forms. One area is building a new operating system with as goal high reliability. Current operating systems are structured as large monolithic lumps of code and a bug in any component, especially in a device driver, can bring down the entire system. The goal of this subproject is to compartmentalize the


40

components of the system so that bugs and security breaches in one compartment cannot spread to other ones. Another area is that of ubiquitous computing, including sensor networks and RFID. Consider a network of sensors deployed in a forest with the intention of detecting fires. A malicious person who looks for an finds a sensor can put a match near it and set off a false alarm. Worse yet, he can do it repeatedly. Using the (software) concept of one-time sensors, we hope to make this problem more manageable. This work is applicable to any sensor network used for safety purposes. Another area of research is RFID security and privacy. RFID tags are being place on many products, on clothes, and passports, and elsewhere. The European bank is thinking of putting them on euro banknotes. Anyone with an easily-available RFID reader can read these tags at a distance. We are looking at ways to guard people's privacy in an RFID-enabled world. Part of our security work focuses on detecting intrusions and blocking malware. We implemented very accurate intrusion detection technology for next-generation honeypots. The system, known as Argos, is based on an hardware emulator which keeps track of what data comes the network and triggers an alert when such data is used in ways that violate the security policy. The system is intended to not just detect the attack, but also generate signatures of the attack. The signatures generated by ourselves and others are used for blocking malicious traffic. For this purpose, we have taken the notion of a distributed firewall to its extreme, whereby an intrusion prevention system is implemented on a programmable network card, close to the wire. The prototype card, known as the CardGuard, may be plugged into an end-user's PC, or, more likely, in the edge router or switch. We are currently developing a more advanced version of CardGuard, known as SafeCard, which applies several different intrusion prevention techniques in low-level hardware. As a result, we expect to offer sophisticated protection at gigabit rates. Finally, we are looking at secure peer-to-peer content distribution networks in which people who buy digital content can be authorized by the owner to resell the content in a secure way. Cluster and Grid Computing Part of this effort is the Ibis project. Ibis is a Java-centric programming environment for writing distributed supercomputing applications. The Ibis project integrates results from earlier projects in this group, including Manta (a Java compiler and fast communication library) and Albatross (which studied parallel programming on wide-area systems, so-called distributed supercomputing). Since 2004, the Ibis project takes place in the context of the BSIK project "Virtual Laboratory for e-Science" (VL-e). The key idea behind Ibis is to benefit from Java's high portability (write once, run anywhere), making it possible to run parallel applications on heterogeneous large-scale systems (Grids). Ibis provides a range of communication primitives (Remote Method Invocation, group communication, divide-and-conquer, replicated objects, message passing) implemented in a portable and efficient fashion. The divide-and-conquer system implemented with Ibis (called Satin) also is fault-tolerant, so it recovers transparently from processor crashes. The core of Ibis has been implemented in pure Java, allowing it to run on any platform that provides a Java Virtual Machine. We have also developed a peer-to-peer called Zorilla that can be used to deploy Ibis (and other) applications on a large-scale system, using locality-aware scheduling. Ibis applications have access to Grid resources (like files or computer nodes) using the Grid Application Toolkit (GAT), for which we have developed a Java-based version within the EC-funded project GridLab. The GAT provides a set of simple APIs for Grid-aware applications, allowing to dynamically bind to proxies that give access to actual resources, services, and protocols. Many large-scale experiments have been performed with Ibis, on the Distributed ASCI Supercomputer (DAS) as well as on a testbed of the EC GridLab project and on the French Grid'5000 system. The experiments have shown that Ibis makes it possible to run a single parallel Java application on a very heterogeneous, European-scale grid and obtain high speedups. For example, we have won an award at the N-Queens Challenge of the Grid Plugtests (Nice, France, Oct. 2005) for using the largest number (961) of CPUs in a single Grid application. Several applications have been developed with Ibis, including automated protein identification (with AMOLF), electromagnetic simulation (JEM3D, with researchers from the French ProActive project), processing of MEG scanner data (with VUmc), and grammar induction (with the University of Amsterdam). In high-speed computing, we also address the issue of processing traffic at truly high link rates (multiple gigabits per second). In this area we are developing the FFPF/Streamline framework. By means of aggressive copy avoidance, minimal context switching, and judicious use of the cache, Streamline significantly improves performance of modern operating systems. In addition, Streamline takes a 'structured operating systems' whereby simple components can be clicked together in straightforward way in order to build complex applications. As a result, the architecture offers more flexibility to the end-


user, and facilitates incorporation of advanced hardware (such as network processor boards, or FPGAs). The framework is sufficiently flexible to allow for distributed traffic processing whereby a high-speed link is split among several nodes (perhaps on several levels) which each perform part of the processing. As the two-level hierarchy easily extends to more levels, the system is inherently scalable. External projects Industrial cooperations have been set up with Philips Research Laboratories (Eindhoven) and Chess embedded Technologies (Haarlem). In addition, we are heavily involved in DevLab, and Dutch consortium of SMEs aiming to apply wireless sensor technology. Efforts continue in the Bsik IShare project, in which the TU-Delft and VU jointly work on the Tribler . Papers in International journals Aline Baggio and Maarten van Steen. Distributed Redirection for the World-Wide Web. Computer Networks 49 (6), page 743-765, December, 2005 Rob V. van Nieuwpoort, Jason Maassen, Thilo Kielmann and Henri E. Bal. Satin: Simple and Efficient Java-based Grid Programming. Scalable Computing: Practice and Experience 6 (3), page 19-32, September, 2005 Spyros Voulgaris, D. Gavidia and Maarten van Steen. CYCLON: Inexpensive Membership Management for Unstructured P2P Overlays. Journal of Network and Systems Management 13 (2), page 197-217, June, 2005 M.L. Boonk, D.R.A. de Groot, F.M.T. Brazier and A. Oskamp. Agent Exclusion on Websites. Proceedings of The 4th Workshop on the Law and Electronic Agents (LEA 2005), page 13-20, 2005 G. Allen, K. Davis, T. Goodale, A. Hutanu, H. Kaiser, T. Kielmann, A. Merzky, Rob V. van Nieuwpoort, A. Reinefeld, Schintke F., T. Schuett, E. Seidel and B Ullmer. The Grid Application Toolkit: Towards Generic and Easy Application Programming Interfaces for the Grid. Proceedings of the IEEE 93 (3), page 534-550, 2005 K. Bicakci, Crispo B. and A. S. Tanenbaum. How to Incorporate Revocation Status Information into the Trust Metrics fro Public-Key Certification. International Journal for Infonomics I (2), page 1-10, 2005 Gabrielle Allen, Kelly Davis, Tom Goodale, Andrei Hutanu, Hartmut Kaiser, Thilo Kielmann, Andre Merzky, Rob van Nieuwpoort, Alexander Reinefeld, Florian Schintke, Thorsten Sch?tt, Ed Seidel, Brygg Ullmer.The Grid Application Toolkit: Towards Generic and Easy Application Programming Interfaces for the Grid. Proceedings of the IEEE, Vol. 93, No. 3, pp. 534-550, March 2005. Ronald Veldema, Ceriel J. H. Jacobs, Rutger F. H. Hofman, Henri E. Bal: "Object combining: a new aggressive optimization for object intensive programs Concurrency and Computation: Practice and Experience", Concurrency & Computation: Practice & Experience, Vol. 17, No. 5-6, pp. 439-464, April/May 2005 Rob V. van Nieuwpoort, Jason Maassen, Gosia Wrzesinska, Rutger Hofman, Ceriel Jacobs, Thilo Kielmann, Henri E. Bal: "Ibis: a Flexible and Efficient Java-based Grid Programming Environment", Concurrency & Computation: Practice & Experience, Vol. 17, No. 7-8, pp. 1079-1107, June/July 2005 Contribution to international conference proceedings Mathijs den Burger, Thilo Kielmann, and Henri Bal: "BalancedMulticasting: High-throughput Communication for Grid Applications", SC'05, Seattle, WA, 12-18 Nov. 2005. C. van Reeuwijk, Rob van Nieuwpoort, and Henri Bal: "Developing JavaGrid Applications with Ibis", Euro-Par 2005, Lisbon, Portugal, August 2005. Olivier Aumage, Rutger Hofman, and Henri E. Bal: "NetIbis: An Efficient and Dynamic Communication System for Heterogeneous Grids", CCGrid05, Cardiff, UK, 9 - 12 May 2005. Gosia Wrzesinskan, Rob van Nieuwpoort, Jason Maassen, and and Henri E. Bal: "Fault-tolerance, Malleability and Migration for Divide-and-Conquer Applications on the Grid", IPDPS 2005, Denver, Colorado, 3-8 April 2005. Spyros Voulgaris and Maarten van Steen. Epidemic-style Management of Semantic Overlays for Content-Based Searching. In: 11th International Conference on Parallel and Distributed Computing (Euro-Par) , Lecture Notes on Computer Science vol. 3648, page 1143-1152, Berlin, September, 2005 Markus Bornemann, Rob V. van Nieuwpoort and Thilo Kielmann. MPJ/Ibis:a Flexible and Efficient Message Passing Platform for Java. In: EuroPVM/MPI 2005, page 217-224, September, 2005


42

H. J. Bos and Huang K. Towards software-based signature detection for intrusion prevention on the network card. In: Eighth International Symposium on Recent Advances in Intrusion Detection (RAID2005), September, 2005 P. Garbacki, D. Epema and Maarten van Steen. A Two-Level Semantic Caching Scheme for Super-Peer Networks. In: 10th Intl Workshop on Web Content Caching and Distribution, September, 2005 Mihai Lucian Cristea, Claudiu Zissulescu, Ed Deprettere and Herbert Bos. FPL-3E: towards language support for reconfigurable packet processing. In: SAMOS V: Embedded Computer Systems: Architectures MOdeling, and Simulation, July, 2005 W. de Bruijn, Herbert Bos and Henri E. Bal. Robust distributed systems: achieving self management through inference. In: 1st International Workshop on Autonomic Communications and Computing, June, 2005 Pietro Mazzoleni, Bruno Crispo, Swaminathan Sivasubramanian and Elisa Bertino. Efficient Integration of Fine-grained Access Control in Large-scale Services. In: Service Computing Conference, page 77-86, June, 2005 Swaminathan Sivasubramanian, Guillaume Pierre and Maarten van Steen. Autonomic Data Placement Strategies for Update-intensive Web Applications. In: International Workshop on Advanced Architectures and Algorithms for Internet Delivery and Applications, June, 2005 Mihai Lucian Cristea, W. de Bruijn and H. J. Bos. FPL-3: towards language support for distributed packet processing. In: IFIP Networking 2005, May, 2005 Swaminathan Sivasubramanian, Gustavo Alonso, Guillaume Pierre and Maarten van Steen. GlobeDB: Autonomic Data Replication for Web Applications. In: 14th International World-Wide Web Conference, page 33-42, New York, NY, May, 2005 B.C. Popescu, J. Sacha, Maarten van Steen, Bruno Crispo, A. S. Tanenbaum and Ihor Kuz. Securely Replicated Web Documents. In: 19th International Parallel & Distributed Processing Symposium, LosAlamitos, CA., April, 2005 Giovanni Russello, M.R.V. Chaudron and Maarten van Steen. Dynamically Adapting Tuple Replication for High Availability in a Shared Data Space. In: 7th Intl Conf. on Coordination Models and Languages, April, 2005 Michal Szymaniak, Guillaume Pierre and Maarten van Steen. Latency-driven Replica Placement. In: Symposium on Applications and the Internet, Los Alamitos, CA., January, 2005 C. Gamage, J. Leiwo, K. Bicakci, Crispo B. and A. S. Tanenbaum. A Cost-Efficient Counter-Intrusion Scheme for One-Time Sensor Networks. In: 2nd International Conference on Intelligent Sensors, Sensor Networks and Information Processing, page 45-50, 2005 K. Bicakci, C. Gamage, Crispo B. and A. S. Tanenbaum. One-Time Sensors: A Novel Concept to Mitigate Node-Capture Attacks. In: 2nd European Workshop on Security and Privacy in Ad-Hoc and Sensors Networks, page 80-90, 2005 K. Bicakci, Crispo B. and A. S. Tanenbaum. Counting Abuses Using Flexible Off-line Credentials. In: 10th Australiasian Conference on Information Security, page 548-559, 2005 Crispo B., Swaminathan Sivasubramanian, Pietro Mazzoleni and Elisa Bertino. P-Hera: Scalable fine-grained access control for P2P. In: 11th International Conference on Parallel and Distributed Systems, page 585-591, 2005 K. Bicakci, Crispo B. and A. S. Tanenbaum. How to Incorporate Revocation Status Information into the Trust Metrics for Publik-Key Certification. In: 20th Annual ACM Symposium on Applied Computing,page 1594-1598, 2005 Herbert Bos, Lai Xu, Kees van Reeuwijk, Mihai Lucian Cristea. Network intrusion prevention on the network card. In: IXA EDU Summit, 2005 Giovanni Russello, M.R.V. Chaudron and Maarten van Steen. Adapting Strategies for Distributing Data in Shared Data Space. In: Intl Symp. on Distributed Objects and Applications, page 1225-1242, 2005 Experimentele software Gobule: a user-centric Content Delivery Network. Java. Solaris/Linux. For description: see above. Cooperation within ASCI The work on Globule and other peer-to-peer related research is done a cooperation between the VU-WI-i and TUD-ti-ki.


2.2.6 Contribution of TUD-EWI-ST-pds Parallel and Distributed Systems Our work in this area covers grids, peer-to-peer systems, wireless sensor networks, and embedded systems. In the area of grids, we have completed the design and implementation of a first version of our KOALA grid scheduler, we have deployed it for general use in the Distributed ASCI Supercomputer, and we have done numerous experiments with it. In addition, we have designed the Grenchmark tool for grid benchmarking. In peer-to-peer systems, we have created a first version of the Tribler p2p client as an extension to Bittorrent with as a final goal TV distribution across the Internet, which includes a novel buddycast algorithm for peer and content discovery, and we have researched semantic-based caching in p2p systems. In the area of wireless sensor networks we participated in the first large-scale (100 nodes) out-door deployment in precision agriculture in The Netherlands in cooperation with Wageningen UR. This pilot showed the large gap between simulations and reality, and produced some detailed RSSI data that we plan to use to redesign our MAC and localization protocols. In the area of embedded systems, we have developed a design of a programming environment and a performance tool for high-performance streaming applications for consumer electronics applications. Cooperative Agent-Based Systems The major effort of the CABS project is directed towards the development of methods and tools for the modelling and implementation of information systems supporting independent, autonomous agents or agent organisations in mutual competition or group-wise cooperation. External Projects Two-level Peer-to-Peer Systems 2003-2007, NWO, EURO 147.000, ASCI partner VU-EW-cs-i. This project aims at exploring the notion of superpeers in p2p systems in order to improve the performance of such systems. Virtual Laboratory for e-Science (VL-e) 2004-2008, EZ (BSIK program), EURO 1.000.000 (TUD-EWI-st-pds part), ASCI partners TUD-EWI-mm-cgcc, VU-EW-cs-I, UvA-FdNWI-caps, UvA-FdNWI-scs.This projects aims at designing and implementing grid technology (schedulers, communication libraries, problem-solving and visualization environments, etc) for virtual laboratories (e.g., for simulations in the sciences) on top of the basic grid fabric. CoreGRID 2004-2008, EU (Network of Excellence), EURO 38.000, ASCI partners VU-EW-cs-i. This network integrates the grid research of 42 universities in Europe. I-SHARE (part of Freeband) 2004-2008, EZ (BSIK program) EURO 800.000, ASCI partners TUD-EWI-mm-ict, VU-EW-cs-i. This projects researches sharing technology in distributed systems, particularly for video. As a research vehicle, we have chosen P2P-TV, a system for sharing live and recorded TV programs of 10,000+ TV channels and web cams among millions of users. SCALP 2004-2008, STW, EURO 220.000, ASCI partners VU-EW-cs-i. This project deals with high-productivity methods for programming parallel systems on a chip. CONSENSUS 2002 – 2006, NWO, Euro 155.000. The CONSENSUS project will develop collaborative algorithms for Wireless Sensor Networks to overcome the limited capabilities of individual sensor nodes. TUD focuses on algorithms for communications and networking, including distributed wireless access, ad-hoc routing protocols, and reliable end-to-end transport. MilSens 2003 – 2007, TNO-FEL (AIO fonds), Euro 160.000. The key objective is to develop an integrated approach to operate large ad-hoc networks of wireless sensor nodes. We will take a practical approach and start from some well-defined application scenarios (to be developed at the start of the project), derive the requirements on the network, and integrate and develop the necessary algorithms. Smart Surroundings 2004 – 2008, EZ (BSIK program), Euro 800.000.


44

The overall mission of the Smart Surroundings project is to investigate, define, develop, and demonstrate the core architectures and frameworks for future ambient systems. TUD focuses on two important aspects at the middleware layer: service discovery and localization. These two are related through the ServiceGRID approach, in which location serves as an index in a database of cached service providers. LOFAR-agro 2004 - 2008, NN (Stichting Samenwerking Noord Nederland), Euro 200.000. The LOFAR-Agro project carries out a pilot project in which ensor nodes will measure the conditions in a potato field; this detailed information (1 reading per 150m2) will be used to improve the advice on how to fight the fungous Phytophtora infestans disease within a crop. RELATE 2005-2008, EU (FP6-IST), EURO 400.000 This project will investigate an approach wherein spontaneously networked mobile objects perform collaborative sensing and communication tocollectively determine their relative positions and spatial arrangement. Contributions to international conference proceedings Feldman, AB ,Gemund, AJC van, & Bos, A . A hybrid approach to hierarchical fault diagnosis. In R Dearden & S Narasimhan (Eds.), Proceedings of DX'05 (pp. 101-106). Pacific Grove, CA, USA: NASA AMES. Pietersma, J,Gemund, AJC van, & Bos, A . A model-based approach to sequential fault diagnosis. In S Karlovic & D Wallhermfechtel (Eds.), Autotestcon 2005 Proceedings (pp. 621-627). Piscataway: IEEE. Garbacki, PJ,Epema, DHJ, & Steen, M van . A two-level semantic caching scheme for super-peer networks. In s.n. (Ed.), Proceedings of the 10th International Workshop on Web Content Caching and Distribution (WCW'05) . Piscataway: IEEE. (TUD) Mors, AW ter,Valk, JM, & Witteveen, C . An event-based task framework for disaster planning and decision support. In B van de Walle & B Carle (Eds.), Proceedings of ISCRAM2005 - Second international ISCRAM conference (pp. 151-153). Brussels, Belgium: KVAB. (TUD) Iosup, A,Garbacki, PJ,Pouwelse, JA, & Epema, DHJ . Analyzing BitTorrent: Three lessons from one-peer level view. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging (pp. 96-103). Delft: ASCI. Ditzel, M, & Langendoen, KG . D3: Data-centric data dissemination in wireless sensor networks. In s.n. (Ed.), European Microwave Week 2005 Conference Proceedings (pp. 185-188). London: Horizon House Publications Ltd. Mors, AW ter, & Witteveen, C . Coordinating non cooperative planning agents: Complexity results. In B Werner (Ed.), Proceedings of the 2005 IEEE/WIC/ACM International Conference on Intelligent Agent Technology (pp. 407-413). Piscataway: IEEE. Wang, J,Reinders, MJT,Lagendijk, RL, & Pouwelse, JA . Distributed collaborative filtering for peer-to-peer file sharing systems. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 - Proceedings of the eleventh annual conference of the advanced school for computing and imaging (pp. 34-40). Delft, The Netherlands: Advanced School for Computing and Imaging (ASCI). Mohamed, HH, & Epema, DHJ . Experiences with the KOALA co-allocating scheduler in multiclusters In s.n. (Ed.), Proceedings of the 5th International Symposium on Cluster Computing and Grid, CCGrid05 (pp. 784-791). Piscataway: IEEE. (TUD) Mohamed, HH, & Epema, DHJ . Experiences with the KOALA co-allocating scheduler in multiclusters. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging (pp. 104-111). Delft: ASCI. (TUD) Haratcherev, IJ,Taal, JR,Langendoen, KG,Lagendijk, RL, & Sips, HJ . Fast 802.11 link adaptation for real-time video streaming by cross-layer signaling. In s.n. (Ed.), Proceedings of IEEE international symposium on circuits and systems (ISCAS 2005) (pp. 3523-3526). Piscataway: IEEE. (TUD) Parker, TEV, & Langendoen, KG . Guesswork: Robust routing in an uncertain world. In P Krishnamurthy (Ed.), Conference proceedings of the 2005 IEEE Conference on Mobile Adhoc and Sensors Systems (pp. 1-9). Piscataway: IEEE. (TUD)


Haratcherev, IJ,Langendoen, KG,Lagendijk, RL, & Sips, HJ . Link adaptation and cross-layer signaling for wireless video-streaming in a shared medium. In s.n. (Ed.), Proceedings of the 2005 International Conference on Wireless Networks, Communications and Mobile Computing (pp. 1-5). Piscataway: IEEE. (TUD) Pouwelse, JA, Slobbe, M van,Wang, J,Reinders, MJT, & Sips, HJ . P2P-based pvr recommendation using friends, taste buddies and superpeers. In M Setten, S McNee, & J Konstan (Eds.), Proceedings of the workshop beyond personalization 2005 - A workshop on the next stage of recommender systems research (IUI2005) (pp. 66-71). s.l.: s.n.. (TUD) Roos, N, & Witteveen, C . Problem solving in a computational society. In HR Arabnia & R Joshua (Eds.), Proceedings of the 2005 International Conference on Artificial Intelligence (pp. 764-770). Las Vegas: CSREA Press. (TUD) Thelen, J, Goense, D, & Langendoen, KG . Radio wave propagation in potato fields. In s.n. (Ed.), Proceedings of the WINMEE, RAWNET and NETCOD 2005 workshops: Measurements, resource allocation and network coding in wireless networks (pp. 1-5). s.l.: ICST. (TUD) Thelen, J, Goense, D, & Langendoen, KG . Radio wave propagation in potato fields. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging (pp. 344-349). Delft: ASCI. (TUD) Wang, J,Reinders, MJT,Lagendijk, RL, & Pouwelse, JA . Self-organizing distributed collaborative filtering. In G Marchionini, A Moffat, J Tait, R Baeza-Yates, & N Ziviani (Eds.), Proceedings of the 28th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 659). New York: ACM. Garbacki, PJ,Epema, DHJ, & Steen, M van . Two-level semantic caches in super-peer networks. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging (pp. 26-33). Delft: ASCI. Wang, J,Reinders, MJT,Pouwelse, JA, & Lagendijk, RL . Wi-Fi Walkman: a wireless handhold that shares and recommends music on peer-to-peer networks. In S Sudharsanan, S Bove, & S Panchanathan (Eds.), Proceedings of SPIE -IS&T Electronic Imaging 2005 Vol. 5683. Proceedings SPIE (pp. 155-163). Bellingham: SPIE. Papers in international Journal Reeuwijk, C van, & Sips, HJ . Adding tuples to Java: a study in lightweight data structures Concurrency and computation-practice & experience, 17(5-6), 423-438. Haratcherev, IJ,Taal, JR,Langendoen, KG,Lagendijk, RL, & Sips, HJ . Automatic IEEE 802.11 rate control for streaming applications Wireless communications & mobile computing, 5(4), 421-437. Halkes, GP, Dam, T van, & Langendoen, KG . Comparing energy-saving MAC protocols for wireless sensor networks Mobile networks & applications, 10(5), 783-791. Mors, AW ter,Valk, JM, & Witteveen, C . Complexity of task coordination for non cooperative planning agents. In M Pechoucek, P Petta, & LZ Varga (Eds.), Multi-agent systems and applications IV Vol. 3690. Lecture notes in artificial intelligence (pp. 600-603). Berlin: Springer. Valk, JM,Weerdt, MM de, & Witteveen, C . Coordination in multi-agent with an application in logistics. In I Vlahavas & D Vrakas (Eds.), Intelligent techniques for planning (pp. 194-224). Hersey: IDEA Group Publishing. Witteveen, C, & Roos, N . Diagnosis of plan execution and the agent. In U Furbach (Ed.), KI 2005: Advances in artificial intelligence Vol. 3698. Lecture notes in artificial intelligence (pp. 161-175). Berlin: Springer. (TUD) Roos, N, & Witteveen, C . Diagnosis of plans and agents. In M Pechoucek, P Petta, & LZ Varga (Eds.), Multi-agent systems and applications IV Vol. 3690. Lecture notes in computer science (pp. 357-366). Berlin: Springer. Pouwelse, JA,Garbacki, PJ,Epema, DHJ, & Sips, HJ . The Bittorrent P2P file-sharing system: Measurement and analysis. Lecture notes in computer science, 3640, 205-216. Mohamed, HH, & Epema, DHJ . The design and implementation of the KOALA co-allocating grid scheduler. Lecture notes in computer science, 3470, 640-650.


46

Verwer, SE,Weerdt, MM de, & Witteveen, C . Timed automata for behavorial pattern recognition. In K Verbeeck, K Tuyls, A Nowé, B Manderick, & B Kuijpers (Eds.), BNAIC 2005 - Proceedings of the seventeenth Belgium-Netherlands conference on artificial intelligence (pp. 291-296). Brussels, Belgium: KVAB. Garbacki, PJ, Biskupski, B, & Bal, H . Transparent fault tolerence for grid applications. Lecture notes in computer science, 3470, 671-680. Goense, D, Thelen, J, & Langendoen, KG . Wireless sensor networks for precise Phytophthora decision support. In JV Stafford (Ed.), Precision agriculture '05 (pp. 573-581). Wageningen: Wageningen Academic Publishers. Experimental software Mohamed, H.H. and Epema, D.H.J.; The KOALA Grid Scheduler 2002-2005, Java, Linux. This scheduler has been designed and implemented for scheduling jobs across multicluster systems and grids. It smain features are its support for co-allocation, i.e., for jobs that can be split up into components across multiple clusters, and its support for load-balancing across clusters. It has been deployed for general use on the DAS in sept. 2005. G.P. Halkes, MAC simulator 2004-2005, OMNeT++, Linux The simulation framework used for our MAC layer publications has been made available for public use in 2004, and was actively supported in 2005. It currently includes models of four energy-efficient protocols for use in Wireless Sensor Networks: S-MAC, T-MAC, B-MAC, and LMAC. Varbanescu, A.L., Sips H.J. and van Gemund A.J.C.; PamSoc 2005-2005, C, Linux PamSoc is a semi-static perforrmance analysis tool for MPSoC systems. It takes as input a program and a parallel machine description and generates a symbolic cost model of an application.

2.2.7 Contribution of TUD-me-ce Arachne The general goal of the project is to investigate novel processor architectures that enables a ubiquitous (i.e., anywhere and anytime) and unobtrusive (i.e., without much user intervention) communication environment. In order to achieve this, we focus on the following (intertwined) research topics: Arachne network processor This entails the specification of the Arachne network processor. In a truly ubiquitous communication environment, the communication devices are not located in a fixed geographical location and therefore may encounter different communication settings during their utilization. For example, moving from a GSM/UMTS network to a Wi-fi network to a Bluetooth network. Instead of incorporating a multitude of heterogeneous and application-specific network processors, we seek to provide a single solution that adapts itself depending on its surroundings in order to lower design complexity and costs of such devices. Herein, we focus on the design of complex hardware units that are able to perform the required functions and on the utilization of reconfigurable hardware technologies, e.g. field-programmable gate arrays (FPGAs). Furthermore, we focus on low-power implementations of the Arachne network processor that are intended to be incorporated in battery-powered handheld devices. Finally, the research in this project also focus on other communication devices other than wireless handheld devices as suggested previously. Examples of such devices include wireless base stations, network routers, network gateways, etc.. Therefore, we investigate the possibility of a single and scalable solution that can be incorporated in these devices. This means covering bandwidths from several kilobits to several tens of Gigabits. Internet and protocol processing This entials the investigation into the functional and temporal requirements of Internet-related applications, such as web-based databases, database mining, content-based processing (e.g., multimedia stream processing), voice-over-IP, videoconferencing, e-learning. In addition, we investigate the requirements associated with the processing of protocols, e.g., IPv4/IPv6, TCP, RTP/RTCP, SIP, etc. and their derivatives that govern and enable the previously mentioned applications on top of the internet. The research performed entails ranging from the modeling and profiling of applications to modeling of the Arachne processor to the design of specialized hardware units. Software and design tools The research performed entails the development of software tools that allow (faster) modeling and profiling of Internet applications and the Arachne network processor with the purpose of investigating the attainable performance and power requirements/consumption. Related research topics include: simulation tools development, benchmarking, application and processor modeling. In addition, special attention must be paid in the integration of these tools as they provide a valuable source of information prior to the definition of the Arachne network processor (targeting a specific network environment or


application set) or to the actual design of specialized hardware units. Finally, design tools must be developed that incorporate the earlier determined profiling information. Molen The MOLEN proposition is that reconfigurable processors, i.e. processors that adapt (dynamically or statically) their microarchitecture to fit application "design requirements", are the answer to the processor (embedded or not) hardware design challenges. To prove the viability of the proposition, we are working on multiple design aspects of (single and multi) processors on a chip using reconfigurable fabric. More specifically our team is working on the following: The MOLEN reconfigurable microcoded processor The main idea for the MOLEN reconfigurable processors is to utilize microcode and custom configured hardware to improve embedded system computing. The reconfigurable hardware execution of code (ranging from a single instruction to a piece of application code) is divided into two logical phases. In the first phase the reconfigurable hardware is being configured. In the second phase the execution of the code is being performed. In both phases, microcode is utilized to perform both the reconfiguration process and the execution of the code. Frequently utilized microcode resides permanently within the fixed part of an on-chip storage facility and non-frequently utilized microcode are paged into the pageable part of the same or another storage facility. The approach is generic, therefore, different applications can utilize the proposed processing capabilities. Our experimentation thus far has involved multimedia operations. In the multimedia experimentation, we investigate processing elements that are capable of performing operations and algorithms found in generating, coding, and displaying multimedia formats, i.e., pictures, video, audio, and graphics. At the current stage, the multimedia processor architecture has targeted multimedia standards including JPEG, MPEG-1/2, MPEG-4, and H.261. Currently, we consider graphic operations and power consumption. We have implemented a Molen processor prototype on the Virtex-II Pro FPGA family from Xilinx Corp. The Virtex-II Pro devices incorporate up to four PowerPC 405 GPP cores, FPGA reconfigurable fabric hardware, dedicated RAM blocks, and dedicated high-speed I/O blocks. In the future of the project we intend to consider lossless compression. Reconfigurable arithmetic and logic processor units The first basic goal is to speed up scientific (mostly vector based) code. Arithmetic (mostly complex to design in hardware) units normally are not present in general purpose processor instruction sets. Such operations include matrix multiplication, sparse matrix operations (such as transpose) etc. They can be implemented in reconfigurable hardware speeding up the execution of scientific programs. A second goal is to design a router and network related reconfigurable hardware. Reconfigurable processor units can be added to general purpose processors for domains (such us switches, networks, packet processing, protocols), that have not been envisioned for the general purpose processor paradigm providing substantial speed-ups. Embedded IP execution units We analyze embedded system computational requirements in order to determine the feasibility of hardwired accelerator units and propose implementations for such units. We have considered JPEG, MPEG-1/2, MPEG-4, H.261, and lossless compression algorithms and we have proposed numerous specialized units including DCT/IDCT, sum of absolute differences (SAD), variable length decoding (VLD), Paeth encoding for portable network graphics (PNG), filters, entropy decoders, repetitive padding units, saturated arithmetic units, accepted quality function (AQF), color space convertors. For our experimentation, we have utilized various FPGA technologies and applied the Molen processor framework to the Philips Trimedia, the IBM's PowerPCs (processors integrated on the Xilinx Virtex-II Pro devices) and the ARM (processors integrated on the Altera Excalibur devices). We will keep on exploring embedded applications for potential hardwired IP units. Memory architecture and implementations Multimedia and embedded processing has specific requirements for memory accesses. For high performance processing, it is required that the memory is accessed in a rectangular manner implying that to be efficient, mechanisms are needed that access memory in a two-dimensional manner. We propose mechanisms for media reconfigurable processors utilizing special addressed memory organizations and an implementation of two-dimensional memory cores that substantially improve the memory performance of the Molen FPGA implemented processor. In addition, because memory may require a significant amount of power, we propose a cache organization that reduces the number of off-chip accesses thus decreasing main memory power consumption of the reconfigurable processor architecture. Compiler and design space exploration tools We have defined a programming paradigm that target the Molen reconfigurable microcoded processor engine and we are developing a backend compiler and a design space exploration toolset. The programming paradigm is based on sequential consistency. It provides mechanisms for parallel and concurrent hardware execution and it is intended (currently) for single program execution. In order to conform to the Molen programming paradigm, an existing compiler has been extended to support the required instruction set and register set extensions. Moreover, a specific mechanism has been developed for passing parameters/results in the case of parallel executions. The compiler and the design space exploration tools are developed in the project The Delft Workbench.


48

Low-Power High-Performance Graphics Architectures We are designing a low-power 2D/3D graphics hardware accelerator for mobile terminals equipped with an ARM processor core. The purpose of using a graphics accelerator is to move some of the graphics-related computations, in particular the rasterization, from the CPU to this dedicated hardware device in order to improve the rendering speed for graphics applications. One important concern for a graphics accelerator meant to be employed in mobile terminals is a low power consumption figure since the most current graphics accelerators are notorious for their high power consumption. Therefore, algorithmic- and circuit-level techniques for low-power graphics need to be studied and evaluated. GraalBench Low-Power Graphics benchmark. The GraalBench is a 3D graphics benchmark suite suitable for 3D graphics on low-power mobile systems, in particular mobile phones. These benchmarks were collected to facilitate our studies on low-power 3D graphics accelerators in the Graal (GRAphics AcceLerator) project. It includes traces of several games as well as virtual reality applications such as 3D museum guides. Applications were selected on the basis of several criteria such as resolution, polygon count, pixel rate, and relevance to mobile devices. For example, 3D FPS games or 3D virtual guides were considered relevant while CAD/CAM applications, such as contained in the Viewperf package, were excluded because it is unlikely that they will be offered on mobile devices (they often have high polygon count and require high resolution). More information and downloads can be found in The GraalBench Benchmark Suite. System on Chip (SoC). Clustered Torus The underlining assumption of this research is that entire systems are migrating onto single chips and that a single chip incorporates numerous heterogeneous IPs. We further assume that the entire system comprises of a network that provides communications among chiplet IPs (processors). In essence, we consider embedded multiprocessor systems and networks on a chip (NoC). We believe that for such systems the interconnection networks have to be regular, expendable, and that they have to provide fast interconnections for some but not all communications between processors. Additionally, networks have to be reliable and avoid lifelock and deadlocks. In our research, we have been proposing a network topology, denoted as the clustered torus, having the properties discussed above and we examine the properties, performance, design issues, and feasibility of such a network. The Delft Java processor We have developed a processor which is a parallel multi-threaded engine optimized for the Java language. Since Java bytecodes are interpreted, the interpretation of the program achieves a performance level that meets or exceeds natively compiled code. The multi-threaded processor architecture is currently being utilized by Sandbridge Technologies to develop cost-effective low-power broadband wireless processor technology. The processor, a 3G digital wireless multi-threaded processor architecture, is intended to meet demands for services such as web browsing, MP3 audio, MPEG-4 video, and video telephony for handheld devices. Contributions to Books T. Hamalainen, A. D. Pimentel, J. Takala, S. Vassiliadis, Embedded computer systems: architectures, modeling and simulation, Springer Verlag, pp. 476, July 2005 Papers in international journals G.K. Kuzmanov, S. Vassiliadis, J.T.J. Eijndhoven, Hardwired MPEG-4 Repetitive Padding, IEEE Transactions on Multimedia, pp. 261--268, April 2005, vol. 7, issue 2 M. Sima, S. D. Cotofana, J.T.J. Eijndhoven, S. Vassiliadis, L.J. Visser, IEEE-Compliant IDCT on FPGA-Augmented TriMedia, Journal of VLSI Signal Processing, pp. 195-212, March 2005, vol. 39, No. 3 Contributions to international conference proceedings G.K. Kuzmanov, G. N. Gaydadjiev, S. Vassiliadis, The Molen Media Processor: Design and Evaluation, Proceedings of the International Workshop on Application Specific Processors, WASP 2005, pp. 26--33, New York Metropolitan Area, USA, September 2005 I. Sourdis, D.N. Pnevmatikatos, S. Wong, S. Vassiliadis, A Reconfigurable Perfect-Hashing Scheme for Packet Inspection, proceedings of 15th International Conference on Field Programmable Logic and Applications (FPL 2005), pp. 644-647, Tampere, Finland, August 2005 G.N. Gaydadjiev, S. Vassiliadis, Flux Caches: What Are They and Are They Useful?, Proceedings of the 5th International Workshop on Computer Systems: Architectures, Modelling, and Simulation (SAMOS 2005), pp. 93-102, Samos, Greece, July 2005, LNCS 3553


E. Moscu Panainte, K. Bertels, S. Vassiliadis, Interprocedural Optimization for Dynamic Hardware Configurations, Proceedings of SAMOS 05, pp. 2-11, Samos, Greece, July 2005, Springer-Verlag Lecture Notes in Computer Science (LNCS) C.J. Glossner, S. Dorward, S. Jinturkar, M. Moudgill, E. Hokenek, M. J. Schulte, S. Vassiliadis, Sandbridge Software Tools, Proceedings of SAMOS 05, pp. 269-278, Samos, Greece, July 2005, Springer-Verlag Lecture Notes in Computer Science (LNCS) E. Moscu Panainte, K. Bertels, S. Vassiliadis, Instruction Scheduling for Dynamic Hardware Configurations, Proceedings of Design, Automation and Test in Europe 2005 (DATE 05), pp. 100-105, Munich, Germany, March 2005 Y. Dou, S. Vassiliadis, G.K. Kuzmanov, G. N. Gaydadjiev, 64-bit Floating-Point FPGA Matrix Multiplication, ACM/SIGDA Thirteenth International Symposium on Field Programmable Gate Arrays (FPGA 2005), pp. 86--95, Monterey, CA, USA, February 2005 2.2.8 Contribution of UvA-FdNWI-csa Microgrids In CSA we are developing new microprocessor architectures that are completely scalable in area, performance and power. These are based on a process of code fragmentation, we call microthreads. Each code fragment is managed in the hardware with minimal and above all scalable structures. Along with the concept of microthreading we are exploring microcontexts and microgrids in order to provide complete solutions to the problems of the highly concurrent microprocessors we expect to see in the near future, i.e. 100s of processors in the next generation and perhaps 100,000 by 2020, by which time CMOS will be at its limit of scaling. System-on-Chip & SoC Design Designers of System-on-Chip (SoC) based embedded systems are typically faced with conflicting design requirements regarding performance, flexibility,, power consumption, and cost. As a result, SoC-based embedded systems often have a heterogeneous system architecture, consisting of components that range from fully programmable processor cores to fully dedicated hardware blocks. Programmable processor technology is used for realizing flexibility, for example to support multiple applications and future extensions, while dedicated hardware is used to optimize designs in time-critical areas and for power and cost minimization. System-level design should be accompanied by a proper methodology for effective and efficient design space exploration. Due to the systems' complexity, it is imperative to have good (performance evaluation) tools for exploring a wide range of design choices, especially during the early design stages where the design space is at its largest. External projects Microgrids NWO, jaartallen : 2005-2009, EURO 250 k Microgrids is an NWO project funded under the Glance program. This project investigates a novel approach to micro-architecture that supports massive on-chip concurrency, which is scalable, flexible and amenable to analysis. It has the potential to provide for the management of on-chip resources (processors etc.) so as to autonomously configure a system for performance, power dissipation or fault tolerance. Artemis STW/Progress,2005-2009, Euro 500K is a cooperation between the University of Amsterdam, Leiden University, Delft University of Technology and Philips Research, studies the design of Network-on-Chip (NoC) based MP-SoCs. Our role in this project constitutes of developing a high-level modeling and simulation framework for system-level design space exploration of these NoC-based MP-SoCs. Papers in International Journals A. D. Pimentel, ``The Artemis Workbench for System-level Performance Evaluation of Embedded Systems’’, in the Int. Journal of Embedded Systems, Vol. 1 (No. 7), 2005. A. D. Pimentel, C. Erbas, and S. Polstra ``A Systematic Approach to Exploring Embedded System Architectures at Multiple Abstraction Levels’’, in IEEE Transactions on Computers, pp. 99-112, Vol. 55 (No. 2), Feb. 2006, Electronic publication: December 2005. K. Bousias, N.M. Hasasneh and C.R. Jesshope, ``Instruction- level parallelism through microthreading - a scalable Approach to chip multiprocessors’’, The Computer Journal, published online in December, 2005.


50

Contributions to International Conference Proceedings C.R. Jesshope, ``Micro-grids - the exploitation of massive on- chip concurrency’’, pp 203-223 (Invited paper, HPC 2004 Cetraro, June 2004), In Grid Computing: A New Frontier of High Performance Computing, 14, pp. 203-223, (ed. L. Grandinetti, Elsevier, Amsterdam, 2005). K. Bousias and C.R. Jesshope, ``The Challenges of Massive On- Chip Concurrency’’, in Proc. of ACSAC, Lecture Notes in Computer Science, Volume 3740, pp157 – 170, 2005. I. Bell, N. Hasasneh, and C.R. Jesshope, ``Microgrids and Micro- contexts: Support Structures for Microthread Scheduling and Synchronisation’’, in Proc. 1st MicroGrid Conference, Amsterdam, July 2005 A. D. Pimentel, ``A Case for Visualization-integrated System- level Design Space Exploration’’, in the Proc. of the 5th Int. Workshop on Embedded Computer Systems: Architectures, MOdeling, and Simulation (SAMOS 2005), pp. 455-464, LNCS, Samos, Greece, July, 2005. Books T. D. Hämäläinen, A. D. Pimentel, J. Takala, and S. Vassiliadis, editors, Proceedings of the 5th Int. Workshop on Embedded Computer Systems: Architectures, MOdeling, and Simulation (SAMOS 2005) , LNCS 3553, Samos, Greece, July, 2005.

2.3 C: Image and Multimedia Sensing, Processing, Interpretation and Visualisation

2.3.1 Contribution of UG-CS-svcg 3D Morphological shape analysis We continued work on multi-scale methods based on mathematical morphology to be used in segmentation, classification and object recognition tasks, in particular morphological hat-transform scale spaces and connected operators. An algorithm was developed to compute Euclidean skeletons of 3D data sets in linear time by the integer medial axis transform. The robust method for solving systems of polynomial equations, as developed last year, has been used to analyse Minkowski-sum based similarity measures of 3D objects. This opens up the way to efficiently calculate this type of similarity measure. Surface and volume visualization In the project ‘Particle systems for image segmentation and visualisation’ work was carried out on a novel, physically-motivated method for surface reconstruction that can recover smooth surfaces from noisy and sparse data sets, without using orientation information. The technique involves a new volumetric technique based on regularised-membrane potentials for aggregating the input sample points and a mesh-smoothing paradigm based on a mass-spring system. Perception-based visualisation A common approach for visualising data sets is to map them to images in which distinct data dimensions are mapped to distinct visual features, such as colour, size and orientation. Many of the end user tasks performed on these images involve a form of visual search. There is evidence for perceptual dependencies when simultaneously presenting multiple features. Such dependencies could potentially affect information visualisations that contain combinations of features for encoding information. Perceptual dependencies in information visualisation were assessed by a complex visual search task. The most relevant findings are that for low-contrast displays colour and size are features that can be used independently to represent information, and feature hierarchies are flexible and accurately controllable by manipulating feature contrasts. Visualisation in functional neuroimaging A review of recent wavelet denoising techniques for medical ultrasound and for (functional) magnetic resonance images was completed and potential applications in the clinical investigations of the brain were discussed. A new approach to the visualisation of time-varying multichannel EEG data was developed, based on an extension of the parallel coordinate method. The method, referred to as tiled parallel coordinates (TPC), can be applied to an arbitrary number of time steps, handling the maximum number of channels currently in use. We also designed and carried out an extensive user evaluation, showing that, for a typical EEG assessment task, data evaluation by the TPC method is faster than by an existing clinical EEG visualisation method, without loss of information.


Gene regulatory network identification and visualization Further work on the {multiple ontology visualisation and exploration, system was carried out, based upon an existing open-source graph visualisation framework. This framework supports several basic modules for input, graph management, transformations, metrics, layout, and interactive display, and allows for integrated network visualisation of genomic, metabolic and proteomic information. A prototype application is under construction, based on the use of open source components for data input, filtering, layout and network visualisation. Together with the Molecular Genetics group (RUG) we also redesigned FIVA (Functional Information Viewer and Analyzer) which is capable of processing information contained in clusters of genes exhibiting similar gene expression patterns, thereby constructing functional profiles. The new version supports multiple view types, and uses focus and context techniques. A usability study has been conducted, and results are being analysed. Computational Geometry and Object Modeling We introduced envelope surfaces, a new class of tangent continuous surfaces based on finite collections of balls. Our results generalise earlier work Edelsbrunner on skin surfaces for modelling macromolecules. Envelope surfaces are more flexible than skin surfaces: they allow for local control with an arbitrary number of degrees of freedom, whereas skin surfaces have only a single global control parameter. Earlier work on certified meshing was extended to the context of excursion sets, where a whole family of isosurfaces can be meshed, with little overhead compared to the cost of meshing a single isosurface. This meshing algorithm is currently being implemented. We also obtained preliminary results on approximating smooth curves by biarc- and conic splines. In particular, we derived the complexity (number of elements) of such a spline in terms of the Hausdorff distance between the curve and the approximating spline. External projects Visualization of regulatory gene networks NWO-CW, 2002-2005, 239 keuro, 1 PhD and 1 Postdoc, part of the project Computational Genomics of Prokaryotes, NWO-BMI program (Biomolecular Informatics). Partners: WU, RU, RUG-Molecular Biology, RUG-Computing Science. Certified Shape Reconstruction NWO-EW, 2005-2009, 90 kEuro, 1 PhD. Interactive morphological and wavelet-based volume processing and visualisation NWO-EW, 2005-2009, 500 kEuro, 2 PhD and 1 Postdoc, part of the NWO program VIEW (Visual Interactive Effective Worlds). Pathway and Network Visualisation Bioinformatics Centre (NBIC), 2005-2008, 175 kEuro (excluding matching of 50% by the University of Groningen), 1 postdoc. Part of the SPX - Project Integrative Bioinformatics of the BioRange program. Algorithms for Complex Shapes with certified topology and numerics (ACS) EU, 2005-2008, 246 kEuro, 1 PhD and 1 postdoc. Vegter is coordinator of this project, in which seven European universities and research institutes and one industrial partner participate. Papers in international journals Muresan, L., Renken, R., Roerdink, J. B. T. M., and Duifhuis; H. Automated correction of spin-history related motion artefacts in fMRI: Simulated and phantom data. IEEE Transactions on Biomedical Engineering 52, 8, 2005, 1450–1460. Roerdink, J. B. T. M.; Morphological pyramids in multiresolution MIP rendering of large volume data: Survey and new results; Journal of Mathematical Imaging and Vision 22, 2/3, 2005, 143–157. Jalba, A. C., Wilkinson, M. H. F., Roerdink, J. B. T. M., Bayer, M. M., and Juggins, S.; Automatic diatom identification using contour analysis by morphological curvature scale spaces. Machine Vision and Applications 16, 4, 2005, 217–228. Contributions to international conference proceedings H. Bekker, E. P. Braad, B. Goldengorin; Selecting the roots of a small system of polynomial equations by tolerance based matching, 4th International Workshop on Efficient and Experimental Algorithms, WEA 2005, LNCS 3503, 610-613, Springer-Verlag. H. Bekker, E. P. Braad, B. Goldengorin; Using bipartite and multidimensional matching to select the roots of a system of polynomial equations, International Conference on Computational Science and its Applications, ICCSA 2005, LNCS 3483, 397-406, Springer-Verlag.


52

Hesselink, W. H., Visser, M., and Roerdink, J. B. T. M. Euclidean skeletons of 3D data sets in linear time by the integer medial axis transform. In Mathematical Morphology: 40 Years On (Proc. 7th Intern. Symp. on Mathematical Morphology, April 18-20), C. Ronse, L. Najman, and E. Decenci‘ere, Eds. Springer, Wien, New York, 2005, pp. 259–268. ten Caat, M., Maurits, N. M., and Roerdink, J. B. T. M. Tiled parallel coordinates for the visualization of time-varying multichannel EEG data. In Proc. Eurographics – IEEE VGTC Symposium on Visualization, June 1-3, Leeds, United Kingdom, K. W. Brodlie, D. J. Duke, and K. I. Joy, Eds. The Eurographics Association, 2005, pp. 61–67. N. G. H. Kruithof and G. Vegter, Meshing skin surfaces with certified topology, Computer Aided Design and Computer Graphics, Hong Kong, China, pag. 287--292, 2005. High-lights Multiscale visualisation based on morphological pyramids We developed a multiresolution approach based on morphological pyramids for multiscale maximum intensity projection (MIP) volume rendering, often used in visualisation of medical angiography data, which allows integrated filtering and fast data exploration (based on reduced pyramid data). Such user interaction may take the form of setting window and level, a common practice in a medical setting, or rendering only thin slices of data, as in sliding thin slab visualisation. Special care is required in choosing the morphological operators to ensure that (i) the MIP operation can be carried out in the transformed domain; (ii) the expand operations are only carried out in the 2D image plane; (iii) detail coefficients can be efficiently coded (only nonzero voxels should be coded, which may save up to 95% of memory for sparse data). Further improvements are possible by using morphological connectivity preserving filters in the construction of the pyramid. The best results so far have been obtained by streaming MIP-splatting, which resorts all detail coefficients in decreasing magnitude and projects only those coefficients needed to attain a user-defined accuracy, see Fig. 1.

(a) (b) (c)

Fig. 1 Multiresolution MIP reconstruction from a 2-level morphological pyramid. (a): full-scale MIP; (b) level-1 MIP; (c) streaming MIP. The absolute errors are 0.676 for (b) and 0.301 for (c). The fraction of nonzero detail coefficients taken into account is 3.4% in both cases (from Roerdink, J. B. T. M.; Morphological pyramids in multiresolution MIP rendering of large volume data: Survey and new results; J. of Mathematical Imaging and Vision 22, 2/3, 2005, 143–157).

2.3.2 Contribution of TU/e-vis Information visualization Our research in information visualization has led to new methods for the visualization of various aspects of software systems, including dynamic behaviour, state spaces and metrics. New results were achieved in our Virtual Reality track, especially concerning interaction and modeling to support interaction. External projects GraphVis 2001-2005, NWO, 100KE This project aims at the visualization of large graphs, with thousands to millions of nodes and edges, for applications such as software engineering and web visualization. Solutions are sought in interaction, clustering and above all, effective visual representations. SMARTER 2002-2007, NWO, 572 KE The aim of the project is to develop new methods and techniques for the simulation of anisotropic turbulent transport. These new methods are based on Local Defect Correction (LDC). This project is a collaboration between four research groups at TU/e. Our contribution concerns the development of new visualization methods.


VOLTS 2004-2008, NWO, 315 KE The aim is to develop new methods to gain insight in large state transition graphs, such as produced when formal methods, i.e. process algebra, is applied to model dynamic systems. The project is a collaboration with the group Design and Analysis of Systems at TU/e. MULTISKEL 2005-2008, NWO, 110KE The aim is to develop new methods for the skeletonization of high-dimensional data-sets. RECONSTRUCTOR 2005-2008, NWO, 1,290 KE Development of new methods for the analysis and understanding of large software systems. Collaboration with the Software Engineering group of prof. dr. A. van Deursen, TU Delft. Doctoral degrees Ham, F.J.J. van; November 21 2005; Interactive visualization of large graphs.TU Eindhoven, Eindhoven 2005, iv+189 pp., ISBN 90-386-0704-0, Contributions to Books Telea, A.C.; An open architecture for visual reverse engineering. K. Khan & Y. Zhang (eds.); Managing Corporate Information Systems Evolution and Maintenance. Hershey PA, USA, 2005, Idea Group Publishing, pp. 211-227. Frasincar, F., Telea, A.C. & Houben, G.J.; Adapting Graph Visualization Techniques for the Visualization of RDF Data. Geroimenko, V. and Chen, C. (eds.); Visualizing the Semantic Web, 2nd edition, Springer, ISBN: 1-85233-976-4, pp. 99-116. Papers in international journals Pranovich, S., Achten, H.H., Vries, B. de, and Wijk, J.J. van; Structural Sketcher: representing and applying well-structured graphic representations in early design. International Journal of Architectural Computing, volume 3, 2005, (1), pp. 75-92. Maubach, J. & Telea, A.; The NumLab Numerical Laboratory for Computation and Visualisation. Computing and Visualisation in Science, volume 8, 2005, 1, pp. 1-17. Pretorius, A.J.; Visual analysis for ontology engineering. Journal of Visual Languages and Computing, volume 16, 2005, 4, pp. 359-381. Rushmeijer, H., Wijk, J.J. van, & Turk, G.; Guest editors' introduction: Special section on IEEE Visualization. IEEE Transactions on Visualization and Computer Graphics, volume 11, 2005, 4, pp. 353-354. Contributions to international conference proceedings Broersen, A., & Liere, R. van; Transfer functions for imaging spectroscopy data using principal component analysis. Eurographics/IEEE VGTC Symposium on Visualization (EuroVis 2005), Leeds, UK, June 1-3, 2005, In Proceedings Eurographics/IEEE VGTC Symposium on Visualization, D. Fellner & S. Spencer (eds.), Aire-la-Ville, Switzerland, 2005: Eurographics Association, pp. 117-123. Broersen, A., Liere, R. van & Heeren, R.M.A. (2005); Comparing three PCA-based Methods for the 3D Visualization of Imaging Spectroscopy Data. Proceedings Visualization, Imaging, & Image Processing (VIIP 2005), J.J. Villanueva (ed.), Benidorm, Spain, September 7-9, 2005, IASTED, ACTA Press, pp. 540-545. Holten, D.H.R., Vliegen, R., & Wijk, J.J. van (2005); Visual realism for the visualization of software metrics. VISSOFT 2005, Budapest, Hungary, September 25, 2005, Proceedings 3rd IEEE International Workshop on Visualizing Software for Understanding and Analysis, S. Ducasse & et al. (eds.), IEEE Computer Society, pp. 27-32. Huizing, C., Kuiper, R., Punter, H.T., & Serebrenik, A.; Looking for stability. DDoPS-05, San Diego CA, USA, June 27-28, 2005, Proceedings of Development and Deployment of Product Software 2005, P. Dey, M. Amin, & S. Brinkkemper (eds.), 2005, pp. 291-304. Huizing, C., Kuiper, R., Punter, H.T., & Serebrenik, A.; Looking for stability. 6th Net.Object Days, Erfurt, Germany, September 19-22, 2005B. Proceedings 6th Net.Object Days, Franczyk & et al. (eds.), 2005, pp. 575-585. Kruszynski, K.J, Liere, R. van & Kaandorp, J. ; Quantifying Differences in Skeletonization Algorithms: A Case Study. Proceedings Visualization, Imaging, and Image Processing (VIIP 2005), J. J. Villanueva (ed.), Benidorm, Spain, September 7-9, 2005, IASTED, ACTA Press, pp. 666-673.


54

Kok, A. & Liere, R. van; A Multimodal Virtual Reality Interface for VTK. Multimodal Interaction for the Visualization of Scientific Data, E.V. Zudilova-Seinstra, T. Adriaansen (eds.). Trento, Italia, October 2005, pp 35-42. Liere, R. van , Kok, A., Martens, J-B, Tienen, M. van; Interacting with Molecular Structures: User Performance versus System Complexity. Proceedings IPT / EGVE 2005, In R. Blach, E. Kjems (eds.), Aalborg, Denmark, October 2005, pp. 147-156. Lommerse, G., Nossin, F., Voinea, S.L., & Telea, A.C.; The Visual Code Navigator: an interactive toolset for source code investigation. IEEE Information Visualization 2005, Minneapolis MN, USA, October 23-25, 2005, Proceedings IEEE InfoVis 2005, IEEE Computer Society, pp. 24-31. Middelkoop, R., Huizing, C., Kuiper, R. & Luit, E.; Cooperation-Based Invariants for OO Languages. Proceedings of the International Workshop on Formal Aspects of Component Software (FACS'05), Luis Barbosa and Zhiming Liu (eds), Macau, October 24-25, 2005, number 333 in UNU-IIST Reports, P.O.Box 3058, Macau, pp. 85-99. Moberts, B., Vilanova, A., & Wijk, J.J. van; Evaluation of fiber clustering methods for diffusion tensor imaging. IEEE Visualization 2005, Minneapolis MN, USA, October 23-28, 2005, Proceedings IEEE Visualization 2005, C. Silva, E. Gröller, & H. Rushmeier (Eds.), pp. 65-72. Pretorius, A.J., & Wijk, J.J. van; Multidimensional visualization of transition systems. 9th International Conference on Information Visualisation IV'05, London, UK, July 6-8, 2005, Proceedings 9th International Conference on Information, IEEE Computer Society, pp. 323-328. Rhijn, A. van, Liere, R. van & Mulder, J.; An Analysis of Orientation Prediction and Filtering Methods for VR/AR. IEEE Virtual Reality 2005, Bonn, Germany, March 2005, Proceedings IEEE Virtual Reality 2005, B. Froehlich, S. Julier, H. Takemura (eds.), 2005, pp. 67—74. Sillanpää, M., & Telea, A.C. (2005); Demonstration of the SoftVision software visualization framework. CSMR 2004, Tampere, Finland, March 24-26, 2004, Tools for Software Maintenance and Reengineering (Proceedings 8th European Conference on Software Maintenance and Reengineering, CSMR 2004), M. Di Penta & M. Harsu (eds.), pp. 88-108. Schindler, E., Kok, A. J.F. & Terken, J.M.B. ; Evaluation of Input Modalities for Interaction Tasks Supporting 3D Object Manipulation. Proceedings Multimodal Interaction for the Visualization of Scientific Data, E.V. Zudilova-Seinstra & T. Adriaansen (eds.), Trento, Italia, October 2005, pp. 35-42. Telea, A.C., & Voinea, S.L.; Interactive visual mechanisms for exploring source code evolution. VISSOFT 2005, Budapest, Hungary, September 25, 2005, Proceedings 3rd IEEE International Workshop on Visualizing Software for Understanding and Analysis, S. Ducasse & et al. (eds.), IEEE Computer Society, pp. 52-57. Termeer, M., Lange, C.F.J., Telea, A.C., & Chaudron, M.R.V.; Visual exploration of combined architectural and metric information. VISSOFT 2005, Budapest, Hungary, September 25, 2005, Proceedings 3rd IEEE International Workshop on Visualizing Software for Understanding and Analysis, S. Ducasse & et al. (eds.), IEEE Computer Society, pp. 21-26. Voinea, S.L., Telea, A.C., & Wijk, J.J. van; CVSscan : visualization of code evolution. SoftVis'05, Saint Louis MO, May 14-15, 2005, Proceedings 2005 ACM Symposium on Software Visualization, New York NY, USA: ACM Press, pp. 47-56. Voinea, S.L., Telea, A.C., & Chaudron, M.R.V.; Version-centric visualization of code evolution. EuroVis 2005, Leeds, UK, June 1-3, 2005, Proceedings Eurographics/IEEE VGTC Symposium on Visualization, D. Fellner & S. Spencer (Eds.), Aire-la-Ville, Switzerland: Eurographics Association, pp. 223-230,331. Voinea, S.L., & Telea, A.C.; Visual assessment techniques for component-based framework evolution. Euromicro-SEAA'05, Porto, Portugal, August 30-September 3, 2005, Proceedings of the 31st Euromicro Conference on Software Engineering and Advanced Applications, pp. 168-179. Wijk, J.J. van; The value of visualization. IEEE Visualization 2005, Minneapolis MN, USA, October 23-28, 2005, Proceedings IEEE Visualization 2005, C. Silva, E. Gröller, & H. Rushmeier (eds.), pp. 79-86. Wijk, J.J. van, & Cohen, A.M.; Visualization of the genus of knots. IEEE Visualization 2005, Minneapolis MN, USA, October 23-28, 2005, Proceedings IEEE Visualization 2005, C. Silva, E. Gröller, & H. Rushmeier (eds.), pp. 567-574).


Experimental software Voinea, L., Telea, A.C., G., Nossin, F.; Visual Code Navigator. 2004-2006 C++, MS Windows The Visual Code Navigator is an open toolset for software visualization. The tools implement various visual techniques for getting interactive insight into large software repositories. See http://www.win.tue.nl/~lvoinea/VCN.html Wijk, J.J. van; SeifertView. 2004-2005, Delphi, MS-Windows. SeifertView enables a user to define and visualize Seifert surfaces: Oriented surfaces that are bound by a knot or link. See www.win.tue.nl/~vanwijk/seifertview Cooperation within ASCI Collaboration with BioMedical Image Analysis group, TU/e Moberts, B., Vilanova, A., & Wijk, J.J. van; Evaluation of fiber clustering methods for diffusion tensor imaging. IEEE Visualization 2005, Minneapolis MN, USA, October 23-28, 2005, Proceedings IEEE Visualization 2005, C. Silva, E. Gröller, & H. Rushmeier (Eds.), pp. 65-72.

2.3.3 Contribution of TUE-EE-dmes Video processing: algorithms and architectures The focus of our video processing research is targeted on high-end consumer applications. The current activities fall into two categories, format conversion and video enhancement. The emphasis of both research activities is shifting with the start of new PhD-students. The format conversion activities have so far been focused on improving the cost/performance ratio, through algorithm-architecture co-design of a motion estimation/compensation unit for picture rate up-conversion. This year the emphasis will shift to performance enhancement of the motion estimation unit. The study has led to initial proof that the algorithm-architecture co-design methodology, for motion estimation, yields up to an order of magnitude reduction in area and power dissipation compared to mapping of the algorithms to a fixed though flexible architecture. In 2004, the design was extended to include picture rate up-conversion and in 2005 the algorithm/architecture co-design of an advanced de-interlacing method based on generalized sampling and directional interpolation was completed. Finally a start has been made extending the design with an extremely flexible video enhancement module applying structure-controlled LMS-filters that emerged from the research in the video enhancement area. The focus of the video enhancement activities shifts from resolution up-conversion, necessary since modern displays have more pixels than the video signal contains, to the enhancement of compressed video. The study has led to a thorough benchmarking of the available video-enhancement algorithms, including classification-based filters and neural networks. Furthermore various innovative algorithms, for de-interlacing, coding artefact reduction and colour enhancement, resulted from this work. Particularly, the gained insight of classified filters was found to enable substitution of time-consuming heuristic optimization procedures by classified mixing. The new research, aimed at compressed video enhancement, continues in this direction and has already resulted in a very effective algorithm capable of boosting fine details while suppressing coding artefacts. In subjective tests this, purely spatial, method outperformed competing designs of much higher complexity. An invention disclosure has been submitted and papers are to follow. External projects Video Enhancement 2001-2005, Philips research, 180 k€ The purpose of video enhancement is to improve the subjective picture quality. Conventional focus in this field was mainly on sharpness, contrast, and colour reproduction improvement, as well as on noise reduction. The purpose of this project is to evaluate what has been achieved in this area, and to propose an optimal technique of SD (standard-definition video) to HD (high-definition video) conversion. Design of systems on silicon for video enhancement and format conversion 2002-2006, Philips Research, 180 k€The goal of this research project is the design of a system on silicon for video enhancement and format conversion, aiming at the joint optimisation of algorithms and architecture for a single chip video converter/enhancer. Fields of interest are motion estimation and compensation, picture rate conversion, de-interlacing, resolution enhancement, noise reduction, power optimisation, VLSI design. Video coding artefact reduction 2005-2009, Philips research, 180 k€ The aim of this project is to gather and evaluate what has been achieved (benchmarking) in the area of coding artifact reduction and to propose (synthesize) some optimal techniques for coding artifact reduction, taking into account constraints of implementation in current technologies and keeping in mind the developments in the area of digitally transmitted video.


56

Papers in international journals Beric, A.; Haan, G. de; Sethuraman, R.; Meerbergen, J.L. van:An Efficient Picture-Rate Up-Convertor. Journal of VLSI Signal Processing 2, june 2005. Peters, H.; Meerbergen, J.L. van; Haan, G. de; Sethuraman, R.; Beric, A.; Meeuwissen, P.; Balakrishnan, S.; Alba Pinto, C.A.; Kruijtzer, W.; Ernst, F.; Alkadi, G.:Application specific instruction-set processor template for motion estimation in video applications. IEEE Transactions on Circuits and Systems for Video Techniques 15, nr. 4, 2005, pp. 508-527. Contributions to international conference proceedings Beric, A.; Haan, G. de; Sethuraman, R.; Meerbergen, J.L. van: Algorithm/Architecture Codesign of the Generalized Sampling Theorem Based De-Interlacer. Proceedings of IEEE/ISCAS Conference, May 2005, 2005, pp. 2943-2946. Beric, A.; Sethuraman, R.; Meerbergen, J.L. van; Haan, G. de: Memory-Centric Motion Estimator. Proceedings IEEE Conference on VLSI Design, January 2005, pp. 816-819. Ciuhu, C.; Haan, G. de: Motion Estimation on Interlaced Video. Proceedings of SPIE, Image and Video Communications and Processing, January 2005. Heesch, F.; Klompenhouwer, M.A.; Haan, G. de: Masking Noise in Up-scaled Video on Large Displays. IEEE Digest of the ICCE'05, Session 11.3, paper 4, 2005. Hu, H.; Hofman, M.; Haan, G. de: Content-adaptive neural filters for image interpolation using pixel classification. Proceedings of SPIE, Applications of Neural Networks and Machine Learning in Image Processing IX, January 2005. Zhao, M.; Ciuhu, C.; Haan, G. de: Classification based data mixing for hybrid de-interlacing techniques. Proceedings of the 13th European Signal Processing Conference, EUSIPCO 2005. (pdf-file, 348 kB), 2005. Zhao, M.; Haan, G. de: Content adaptive vertical temporal filtering for de-interlacing. Proceedings of the 9th INternational Symposium on Consumer Electronics, 2005 ISCE 2005 (pdf-file 493 kB), 2005, pp. 69-73. Zhao, M.; Haan, G. de: Subjective evaluation of de-interlacing techniques. Proceedings of SPIE, Image and Video Communications and Processing, January 2005. Experimental software Zhao, M; Resolution Up-Conversion 2003 - 2005, C++/Windows XP. Content-adaptive vertical temporal de-interlacing: De-interlacing of video data using adaptive vertical temporal filters. Classification based data mixing for hybrid de-interlacing techniques: based on classification and Least Mean Square filtering, we can achieve better data mixing results for hybrid de-interlacing. Subjective optimal Least Mean Square filtering for up-scaling: Using the inverse relationship between sharpen and blur, we designed content adaptive up-scaling techniques with build-in sharpness enhancement, thus improving the perceived up-scaling image quality. Experimental hardware Beric, A; Sethuraman, R; A motion compensated picture-rate up-converter 2003 – 2005 This design concerns a VLIW-based ASIP for motion estimation used in picture-rate up-conversion. The ASIP meets low-power and low-cost requirements apart from providing flexibility for the application domain. It consumes 27 mW and takes an area of 1.1 mm2 in 0.13 µm technology for delivering motion estimation functionality for standard definition (SD) sequences at 140fps. The designed hardware was synthesized and tested up to the pre-layout netlist level. Patent Haan, G. de: A unit for and method of image scaling. Philips, EP1547378 A1, 29 June 2005. Haan, G. de: A unit for and method of image conversion. Philips, EP1554874 A1, 20 July 2005. Haan, G. de: A unit for and method of image conversion. Philips, EP1565878, 24 August 2005. Haan, G. de: Gamma correction. Philips, EP1570651 A1, 7 September 2005. Haan, G. de: Method for image scaling. Philips, EP1540593, 15 June 2005. Haan, G. de; Bellers, E.B.: Chrominance signal interpolation. Philips, US6498609, 24 December 2005.


Haan, G. de; Biezen, P.W.A.C.; Wittebrood, R.B: Image processing unit with fall-back. Philips, EP1557037, 27 July 2005. Haan, G. de; Klompenhouwer, M.A.: Anti motion blur display. Philips, US6930676, 16 August 2005. Haan, G. de; Klompenhouwer, M.A.: Method of and unit for processing images. Philips, US6501446, 31 December 2005. Haan, G. de; Klompenhouwer, M.A.; Velthoven, L.J.: A unit for and method of sharpness enhancement. Philips, EP1512120, 9 March 2005. Haan, G. de; Pelagotti, A.: Problem area location in an image signal. Philips, EP1048170, 11 February 2005. Haan, G. de; Wittebrood, R.B: Recognizing film and video objects occurring in parallel in single television signal fields. Philips, US6937655, 30 August 2005. Klompenhouwer, M.A.; Haan, G. de; Velthoven, L.J.: Unit and method of calculating a sharpened edge. Philips, EP1512120, 9 March 2005. Riemens, A.K.; Dommisse, A.; Haan, G. de: Recognizing film and video occurring in parallel in television fields. Philips, EP1574055 A1, 14 September 2005. Velthoven, L.J.; Klompenhouwer, M.A.; Haan, G. de: Edge dependent motion blur reduction. Philips, EP1509881, 26 January 2005. Wesenbeeck, J.M.A. van; Haan, G. de; Klompenhouwer, M.A.; Krijn, M.P.C.M.: CRT with enhanced vertical resolution. Philips, EP1540689, 15 June 2005. Wittebrood, R.B; Haan, G. de: Unit for and method of estimating a motion vector. Philips, EP1514241, 9 March 2005. Wittebrood, R.B; Haan, G. de: Unit for and method of estimating a motion vector. Philips, EP1514242, 16 March 2005. Wittebrood, R.B; Haan, G. de: Image segmentation using template prediction. Philips, EP1565879 A2, 24 August 2005.

2.3.4 Contribution of TUD-mm-ict Pattern Recognition A typical human ability is the recognition of patterns in the world around us. It constitutes the basis of each natural science: the laws of physics, the description of species in biology or the analysis of human behavior; they are all based on seeing patterns. Also in daily life pattern recognition plays an import role: reading texts, identifying people, retrieving objects or finding the way in a city. Once patterns are established, learned from some examples or from a teacher, we are able to classify new objects or phenomena into a class of known patterns. The study of automatic pattern recognition has two sides, one purely fundamentally scientific and one applied. By trying to build a system with pattern recognition capabilities more will become clear about the human ability to learn, recognize and classify. At the same time, systems are constructed that may be applied in various areas of science as well as in many places in society to assist human decision making. In our research, both aspects are treated. There are two projects focusing on the foundations of pattern recognition: representation and generalization, in which new ways of describing objects and learning from examples are studied. In addition, there are several applied projects focusing on the recognition of spectra. Like images and time signals, spectra constitute a pattern domain with an unsolved problem: how to represent patterns in their entirety, when only a set of samples (pixels or bins) is available. The research makes clear that automatic pattern recognition systems may successfully be applied in several places but that an understanding of the human ability of recognizing patterns is still in its early days. Bioinformatics Knowledge discovery and machine learning play an increasingly important role in biomolecular science, now that micro-array technology reaches maturity and genomes of different organisms are being published with increasing frequency. As a consequence, massive amounts of data are being (and will be) produced that contain important facts about the "Book of Life" of living organisms. These measurements have opened new avenues to discover novel functionalities within cells (e.g. function of genes and gene pathways). Although molecular biologists are extremely enthusiastic about these developments, they are confronted with the availability of enormous amounts of data about a very complex (many genes, hence many parameters) and heterogeneous (many elements such as genes, proteins metabolites etc.) system. Traditional ways to study biological phenomena in a gene-by-


58

gene approach (reductionism) are no longer adequate. Instead, the cell should be studied as a network of complex interactions. This systems view, coupled with the integration of already available knowledge, defines the core of our approach to biomolecular knowledge discovery. External projects ALL-AGE ("Analysis of gene expression profiles of children with acute lymphomatic leukaemia by bioinformatics") 2004-2008, EUR, 280 k€. The ALL-AGE project is a co-operation with the Department of Immunology of the Erasmus University Medical Center in Rotterdam. Jointly, we apply, evaluate and develop techniques for the analysis of microarray data and other molecular measurements. Applications include: basis molecular biology, molecular diagnosis and gene therapy. Improving endoscopic detection of lung cancer using autofluorescence spectroscopy analyzed by a neural network 2001-2005, STW, 100k€.This project, a cooperation with the Daniel den Hoed Hospital in Rotterdam, in which we participated by studying the data-analysis. We found that by combining a set of spectral classifiers, each based on a different excitation wavelength a significant improvement of the recognition performance may be reached. However, the size of the dataset collected over the years is still too small for an advanced development of classifiers. The project has been finished in 2005. The analysis of spatial structures in hyperspectral images 2003-2007, STW, 470 k€.The construction of an initial toolbox for hyperspectral image recognition was finalized. We are now focusing on applications in various areas like agriculture, mining, paper production and infra-red imaging for microscopic particle analysis. On a more fundamental level we study procedures fo the selection of subbands, good for classification. One-Class Classifiers 2001-2005, NWO 250 k€. In cooperation with some other projects like the characterization of interstitial lung disease and mineral spectral analysis a set of advanced applications has been studied, resulting in a better understanding of performance estimation of one-class classifiers. In addition we studied techniques for active and semi-supervised learning. The project has been finished in 2005 and concluded by a PhD thesis, defended in 2006. Proximity-based representations for pattern learning 2004-2008, NWO, 370 k€. The essential difference between proximity representations and kernel approached has been studied and resulted in understanding of the possible advantage of non-Euclidean object representations. Further the construction of augmented representation spaces and the application for invariant representations has been studied. Computer-aided detection and characterization of interstitial lung disease 2004-2008, STW, 200 k€.This project started as a cooperation with the Image Science Institute in Utrecht. We intend to make use of the one-class classifier technique and study the use of spatial data-connectivity for improved recognition. Mineral spectral analysis 2003-2005, DeBeers, 90 k€ This project aims at a robust spectral recognition of multi-class minerals for changing class distributions. Extensive experiments have been performed in which many classification techniques have been evaluated. Unknown priors, novelty detection and changing class distributions are the challenges for this real world application. ITB, "Innovative therapies for bone recovery" 2004-2007, Senter/Organon, 290 k€. Due to the continuous increase of people's life expectancy, there is an urgent need for effective therapies for the recovery and prevention of bone loss due to osteoporosis. In this project, the newest developments in genomics are integrated with the latest techniques in pathway discovery and pattern recognition. This will lead to a faster and novel approach to discover and functionally characterize genes involved in bone development. REGNET "Unravelling the hierarchy of transcriptional regulation in chemostat cultures of Saccharomyces cerevisiae: a bioinformatics approach", 2004-2008, NGI/Kluyver Centre for Genomics of Industrial Fermentation, 230 k€.The Kluyver Center generates large transcriptome datasets of Saccharomyces cerevisiae grown under a wide range of carefully defined, industrially relevant, cultivation conditions in steady-state chemostat cultures. These different environmental conditions lead to widely differing physiological responses. The general objective of this project is to implement and develop bioinformatics approaches to unravel the underlying transcriptional regulation networks. Special emphasis will be placed on the hierarchy of transcriptional regulation, i.e., the quantitative analysis of how the (simultaneous) action of different transcription factors acting at the same promoter region determine the transcriptional response of a promoter. Analysis of regulatory networks in Saccharomyces Cerevisiae 2004-2008, NGI/Kluyver Centre for Genomics of Industrial Fermentation, 150k€.


The overall aim of this project is to investigate genome-wide regulatory circuitry (transcription and signal transduction), by concurrently developing and performing bioinformatic analyses of various genome-scale data sets: expression microarray data; chromosome localization microarray data; protein-protein interaction data; high throughout phenotype data; sequence data as well as data currently stored in databases. METNET " Data-driven discovery of metabolic and regulatory network structure", 2004-2008, Delft Research Center Life Science and Technology, 210 k€. This project is collaboration with the Bioprocess Technology Group at the TUDelft that studies mathematical techniques to accurately model metabolic reactions using metabolic control analysis when the structure of networks of such reactions is assumed to be known. In this project, tools developed for genetic network discovery will be adapted and extended to allow ab initio metabolic network structure discovery, facilitating detailed modeling by the Bioprocess Technology Group. These models will subsequently be extended to incorporate regulatory mechanisms at the genomic and proteomic level. Such models will allow targeted interference at the genomic level to obtain certain desired metabolic behaviour. BioRange-SP121, “Unifying framework for data-driven pathway discovery” 2004-2009, NBIC-BioRange, 350 k€ Together with the Radboud University Nijmegen, the University Medical Center Utrecht and the Dutch Cancer institute we study the integration of various genomic sources to arrive at novel pathways. In particular we look at the development of a unifying framework to integrate this data and at the analysis of insertion sites in retroviral insertional mutagenesis screens and relate them to onco-genetic events. BioRange-SP132, “Developing clinical predictors based on high-dimensional genomics data, pathway information and directed experimentation” NBIC-BioRange, 350 k€. Together with the Leiden University Medical Center and the Dutch Cancer institute this project studies data-driven approaches based on genomics data to construct predictors for diagnosis and prognosis in human disease. We will develop statistical techniques which regulate the gene selection strategy such that it takes particular known relationships between genes, defining e.g. a particular pathway, into account. In addition, we explore how predictors based on gene expression data can be augmented with prognostic predictors derived from proteomics data originating from the same patient series. BioRange-SPHGL Bioinformatics, NBIC-BioRange, 450 k€. This project concerns funding by NBIC-BioRange for the chair of Prof. Reinders in Bioinformatics. CANCER "Molecular Classification of Cancer", 2002-2007, NKI, 300 k€. This project is conducted in close collaboration with the Dutch Cancer Institute (NKI). Central to the program lies the prediction of prognosis and optimal treatment choice for cancer patients, which is dependent on correct disease classification. Our goal within this program is to perform computational analysis of all information sources to construct a taxonomy of breast cancer (sub-)types, such that the classes correlate strongly with survival and response to radio- or chemotherapy. Doctoral Degrees Pekalska, E.M.; January 17, 2005 Dissimilarity representations in pattern recognition. Concepts, theory and applications. TU Delft, Delft 2005, 116 pp., ISBN 90-9019021-X, external project Books Pekalska, E.M., Duin, R.P.W.; Dissimilarity representation for pattern recognition: foundations & applications. Singapore, 2005, 636 pp., World Scientific, ISBN 981-256-530-2, Series in Machine Perception and Artificial Intelligence, vol. 64, 1st edition. Contributions to Books Duin, R.P.W., Tax, D.M.J.; Statistical Pattern Recognition. Chen, C.H., Wang, P.S.P.; Handbook of Pattern Recognition and Computer Vision.Singapore, 2005, 967 pp., World Scientific, ISBN 981-256-105-6, pp. 3-24, 3rd edition. Kranenburg, P. van, Backer, E.; Musical Style Recognition – A Quantitative Approach. Chen, C.H., Wang, P.S.P. (eds.); Handbook of Pattern Recognition and Computer Vision, 3rd ed. Singapore, 2005, 967 pp., World Scientific Press, ISBN 981-256-105-6, pp. 583-600. Papers in international journals Bard, M.P.L., Amelink, A., Skurichina, M., Bakker, M. den, Burgers, S.A., Meerbeeck, J.P. van, Duin, R.P.W., Aerts, J.G.J.V., Hoogsteden, H.C., Sterenborg, H.J.C.M.; Improving the specificity of fluorescence bronchoscopy for the analysis of


60

neoplastic lesions of the bronchial tree by combination with optical spectroscopy: preliminary communication. Lung Cancer, Elsevier, ISSN 0169-5002, vol. 47, no. 1, pp. 41-47. Beers, E.H. van, Welsem, T. van, Wessels, L.F.A., Li, Y., Oldenburg, R.A., Devilee, P., Cornelisse, C.J., Verhoef, S., Hogervorst, F.B.L., Veer, L.J. van 't, Nederlof, P.M.; Comparative genomic hybridization profiles in human BRCA1 and BRCA2 breast tumors highlight differential sets of genomic aberrations. Cancer Research, ISSN 0008-5472, vol. 65, no. 3, pp. 822-827 Dik, W.A., Pike-Overzet, K., Weerkamp, F., Ridder, D. de, Haas, E.F.E. de, Baert, M.R.M., Spek, P. van der, Koster, E.E.L., Reinders, M.J.T., Dongen, J.J.M. van, Langerak, A.W., Staal, F.J.T.; New insights on human T cell development by quantitative T cell receptor gene rearrangement studies and gene expression profiling. The Journal of Experimental Medicine, Rockefeller University Press, ISSN 0022-1007, vol. 201, no. 11, pp. 1715-1723. Glas, A.M., Kersten, M.J., Delahaye, L.J.M.J., Witteveen, A.T., Kibbelaar, R.E., Velds, A., Wessels, L.F.A., Joosten, P., Kerkhoven, R.M., Bernards, R., Krieken, J.H.J.M. van, Kluin, P.M., Veer, L.J. van 't, Jong, D. de; Gene expression profiling in follicular lymphoma to assess clinical agressiveness and to guide the choice of treatment. Blood, The American Society of Hematology, ISSN 0006-4971, vol. 105, no. 1, pp. 301-307. Loog, M., Ginneken, B. van, Duin, R.P.W.; Dimensionality reduction of image features using the canonical contextual correlation projection. Pattern Recogntion, Elsevier, ISSN 0167-8655, vol. 38, no. 12, pp. 2409-2418. Ridder, D. de, Linden, C.E. van der, Schonewille, T., Dik, W.A., Reinders, M.J.T., Dongen, J.J.M. van, Staal, F.J.T.; Purity for clarity: the need for purification of tumor cells in DNA microarray studies. Leukemia, Nature Publishing Group, ISSN 0887-6924, vol. 19, no. 4, pp. 618-627. Roepman, P., Wessels, L.F.A., Kettelarij, N., Kemmeren, P., Miles, A.J., Lijnzaad, P., Tilanus, M.G.J., Koole, R., Hordijk, G.-J., Vliet, P.C. van der, Reinders, M.J.T., Slootweg, P.J., Holstege, F.C.P.; An expression profile for diagnosis of lymph node metastases from primary head and neck squamous cell carcinomas. Nature Genetics, ISSN 1061-4036, vol. 37, no. 2, pp. 182-186. Vaes, B.L.T., Dechering, K.J., Someren, E.P. van, Hendriks, J.M.A., Ven, C.J.J.M. van de, Feijen, A., Mummery, C.L., Reinders, M.J.T., Olijve, W., Zoelen, E.J.J. van, Steegenga, W.T.; Microarray analysis reveals expression regulation of Wnt antagonists in differentiating osteoblasts. Bone, Elsevier, ISSN 8756-3282, vol. 36, no. 5, pp. 803-811. Veenman, C.J., Reinders, M.J.T.; The nearest subclass classifier: a compromise between the nearest meand and nearest neighbor classifier. IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE, ISSN 0162-8828, vol. 27, no. 9, pp. 1417-1429. Veld, D.C.G. de, Skurichina, M., Witjes, M.J.H., Duin, R.P.W., Sterenborg, H.J.C.M., Roodenburg, J.L.N.; Autofluorescence and diffuse reflectance spectroscopy for oral oncology. Lasers in Surgery and Medicine, Wiley-Liss, Inc., ISSN 0196-8092, vol. 36, no. 5, pp. 356-364. Vliet, M.H. van, Kempen, G.M.P. van, Reinders, M.J.T., Ridder, D. de; Computational estimation of the composition of fat/oil mixtures containing interesterification from gas and liquid chromatography data. Journal of the American Oil Chemists' Society, American Oil Chemists' Society, ISSN 0003-021X, vol. 82, no. 10, pp. 707-716. Weigelt, B., Wessels, L.F.A., Bosma, A.J., Glas, A.M., Nuyten, D.S.A., He, Y.D., Dai, H., Peterse, J.L., Veer, L.J. van 't; No common denominator for breast cancer lymph node metastasis. British Journal of Cancer, Nature Publishing Group, ISSN 0007-0920, vol. 93, no. 8, pp. 924-932. Wessels, L.F.A., Reinders, M.J.T., Hart, A.A.M., Veenman, C.J., Dai, H., He, Y.D., Veer, L.J. van 't; A protocol for building and evaluating predictors of disease state based on microarry data. Bioinformatics, Oxford University Press, ISSN 1367-4803, vol. 21, no. 19, pp. 3755-3762 Zelm, M.C. van, Burg, M. van der, Ridder, D. de, Barendregt, B.H., Haas, E.F.E. de, Reinders, M.J.T., Lankester, A.C., Révész, T., Staal, F.J.T., Dongen, J.J.M. van; Ig gene rearrangement steps are initiated in early human precursor B cell subsets and correlate with specific transcription factor expression. The Journal of Immunology, American Association of Immunologists, Inc., ISSN 0022-1767, vol. 175, no. 9, pp. 5912-5922.


Contributions to international conference proceedings) Dongen, J.J.M. van, Velden, V.H.J. van der, Ridder, D. de, Langerak, A.W., Staal, F.J.T.; Laboratory diagnosis of leukemia: can we replace current molecular diagnostics by novel flow cytometry? The 10th Congress of the European Hematology Association, Stockholm, 2-5 Juni 2005, Proceedings of the 10th Congress of the European Hematology Association, pp. 36-40. Duin, R.P.W., Pekalska, E.M.; Open issues in pattern recognition. The 4th International Conference on Computer Recognition Systems, Wroclaw, 22-23 May 2005, Proceedings of the 4th International Conference on Computer Recognition Systems, Kurzynski, M., Puchala, E., Wozniak, M., Zolnierek, A. (eds.), Springer Verlag Berlin, ISBN 3-540-25054-9, pp. 27-42. Juszczak, P., Duin, R.P.W.; Learning from a test set. The 4th International Conference on Computer Recognition Systems, Wroclaw, 22-23 May 2005, Proceedings of the 4th International Conference on Computer Recognition Systems, Kurzynski, M., Puchala, E., Wozniak, M., Zolnierek, A. (eds.), Springer Verlag Berlin, ISBN 3-540-25054-9, pp. 203-210. Landgrebe, T.C.W., Duin, R.P.W.; On Neyman-Pearson optimisation for multiclass classifiers. The Sixteenth Annual Symposium of the Pattern Recognition Association of South Africa, Langebaan, 23-25 November 2005, Proceedings of the Sixteenth Annual Symposium of the Pattern Recognition Association of South Africa, Nicolss, F. (ed.), IAPR, ISBN 0-7992-2264-X, pp. 165-170. Landgrebe, T.C.W., Paclik, P., Tax, D.M.J., Duin, R.P.W.; Optimising two-stage recognition systems. The 6th International Workshop Multiple Classifier Systems, Seaside, CA, USA, 13-15 June 2005, Proceedings of the 6th International Workshop Multiple Classifier Systems / Lecture Notes in Computer Science, vol. 3541, Oza, N.C., Polikar, R., Kittler, J., Roli, F. (eds.), Springer Verlag, Berlin, Heidelberg, ISBN 3-540-26306-3, pp. 206-215. Paclik, P., Duin, R.P.W.; Designing multi-modal classifiers of spectra: a study on industrial sorting application. The 2nd International Spectral Imaging Workshop, Austria, September 2005, Proceedings of the 2nd International Spectral Imaging Workshop, Leitner, R. (ed.), Austrian Computer Society, ISBN 3-85403-194-7, pp. 19-25. Paclik, P., Landgrebe, T.C.W., Tax, D.M.J., Duin, R.P.W.; On deriving the second-stage training set for trainable combiners. The 6th International Workshop Multiple Classifier Systems, Seaside, CA, USA, 13-15 June 2005, Proceedings of the 6th International Workshop Multiple Classifier Systems / Lecture Notes in Computer Science, vol. 3541, Oza, N.C., Polikar, R., Kittler, J., Roli, F. (eds.), Springer Verlag, Berlin, Heidelberg, ISBN 3-540-26306-3, pp. 136-146 Paclik, P., Tax, D.M.J., Verzakov, S., Duin, R.P.W.; Simplifying the model-based classifiers for multi-modal problems in classification of spectra. 33-38. Paclik, P., Verzakov, S., Duin, R.P.W.; Improving the maximum-likelihood co-occurrence classifier: a study on classification of inhomogeneous rock images. The 14th Scandinavian Conference on Image Analysis, Joensuu, Finland, 19-22 June 2006, Proceedings of the 14th Scandinavian Conference on Image Analysis / Lecture Notes in Computer Science, vol. 3540, Kalviainen, H., Parkinen, J., Kaarna, A. (eds.), Springer Verlag Berlin Heidelberg 2005, ISBN 3-540-26161-3, pp. 998-1008. Pekalska, E.M, Harol, A., Lai, C., Duin, R.P.W.; Pairwise selection of features and prototypes. The 4th International Conference on Computer Recognition Systems, Wroclaw, 22-23 May 2005, Proceedings of the 4th International Conference on Computer Recognition Systems, Kurzynski, M., Puchala, E., Wozniak, M., Zolnierek, A. (eds.), Springer Verlag Berlin, ISBN 3-540-25054-9, pp. 271-278. Pekalska, E.M., Duin, R.P.W.; The use of dissimilarities for object recognition. The EOS Conference on Industrial Imaging and Machine Vision, Munich, 13-15 June 2005, Proceedings of the EOS Conference on Industrial Imaging and Machine Vision, SPIE, ISBN 3-00-0163161-1, pp. 50-53. Skurichina, M., Duin, R.P.W.; Combining feature subsets in feature selection. The 6th International Workshop Multiple Classifier Systems, Seaside, CA, USA, 13-15 June 2005, Proceedings of the 6th International Workshop Multiple Classifier Systems / Lecture Notes in Computer Science, vol. 3541, Oza, N.C., Polikar, R., Kittler, J., Roli, F. (eds.), Springer Verlag, Berlin, Heidelberg, ISBN 3-540-26306-3, pp. 165-175. Tax, D.M.J., Duin, R.P.W.; Characterizing one-class datasets. The Sixteenth Annual Symposium of the Pattern Recognition Association of South Africa, Langebaan, 23-25 November 2005, Proceedings of the Sixteenth Annual Symposium of the Pattern Recognition Association of South Africa, Nicolss, F. (ed.), IAPR, ISBN 0-7992-2264-X, pp. 21-26. Veenman, C.J., Tax, D.M.J.; A weighted nearest mean classifier for sparse subspaces. Proceedings of the Conference on Computer Vision and Pattern Recognition, San Diego, CA, USA, 20-26 July 2006, Proceedings of he Conference on


62

Computer Vision and Pattern Recognition, Schmid, C., Soatto, S., Tomasi, C. (eds.), IEEE Computer Society, pp. 6. Verzakov, S., Paclik, P., Duin, R.P.W.; The tangent kernel approach to illumination-robust texture classification. The 14th Scandinavian Conference on Image Analysis, Joensuu, Finland, 19-22 June 2006, Proceedings of the 14th Scandinavian Conference on Image Analysis / Lecture Notes in Computer Science, vol. 3540, Kalviainen, H., Parkinen, J., Kaarna, A. (eds.), Springer Verlag Berlin Heidelberg 2005, ISBN 3-540-26161-3, pp. 1009-1016. Cooperation within ASCI Veenman, C.J., Reinders, M.J.T.; The nearest subclass classifier: a compromise between the nearest meand and nearest neighbor classifier. IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE, ISSN 0162-8828, vol. 27, no. 9, pp. 1417-1429. Veenman, C.J., Tax, D.M.J.; A weighted nearest mean classifier for sparse subspaces. Proceedings of the Conference on Computer Vision and Pattern Recognition, San Diego, CA, USA, 20-26 July 2006, Proceedings of he Conference on Computer Vision and Pattern Recognition, Schmid, C., Soatto, S., Tomasi, C. (eds.), IEEE Computer Society, pp. 6.

2.3.5 Contribution of TUE-BT-bmia Multi-scale computer vision Analysis and exploitation of the deep (i.e. over scale) structure of images. This is biologically inspired, fundamental research into singularities and toppoints in images, to come to a hierarchical representation. A NWO VICI grant was rewarded to dr. Luc Florack (see highlights below). A new class of multi-scale image structure (α-scale-spaces) has been developed. Methods have been designed to exploit the extracted toppoints for image retrieval, and to efficiently reconstruct the image from the toppoints, using new methods based on inner products of Sobolev type. Together with Oce BV we develop a toppoint-based document analysis system for the IOP project “Intelligent Scanners”. Perceptual grouping and contect operators. This biologically inspired research focuses on the exploitation of the multi-scale orientation structure of images. A new class of invertible orientation wavelets has been developed, as well as new theories for stochastic completion fields and efficient steerable tensor voting, with applications for the detection of dim elongated structures (contours, catheters, bloodvessels, electrodes) in images. In collaboration with Philips we improved the detection of cardiac electro-physiology catheters at low-dose fluoroscopy. Diffusion tensor imaging analysis and 3D visualization 3D visualization techniques for MR diffusion tensor imaging (DTI) data have been investigated in an interactive setting, in close collaboration with the TUE Dept. of Computing Science. Clinical studies focus on neonatal brain fibre development, muscle fibre orientations in the heart and skeletal muscle, and the extraction of grouped bundles in DTI. We started programming on modern GPU graphics cards. A NWO VENI grant was rewarded to dr. Anna Vilanova (see highlights below). Methods were developed to automatically extract transfer functions from the data itself, i.e. from LH histograms (fig. 2). Illustrative rendering techniques have been developed for medical 3D volume datasets, in the styleof drawing in classicalanatomy books (fig. 5).

Fig. 2. CT of a tooth using a transfer function based on the LH Histogram. The dentin-air (ochre) and enamel-air (white) boundaries are set to be semitransparent to reveal the inner boundaries. Note that part of the pulp-dentine boundary (red) is identical with the ochre boundary [Sereda et al. 2005]. http://www.bmi2.bmt.tue.nl/image-analysis/Research/Projects/20010901-Visualization/index.html

Clinical image analysis & applications This research theme is in close collaboration with Philips Medical Systems, Best. Automated detection of pulmonary emboli in spiral 3D CT acquisitions of the lung. A computer-aided diagnosis system is developed, based on second order shape and intensity features. A detailed analysis was carried out on the dislocation of highly curved surfaces such as of small bloodvessels.


Segmentation of short-axis cardiac MR using active contours. Geometry determination, simulations and motion predictions of dynamic vessel wall behavior in abdominal aortic aneurysms (fig. 3). We collaborate with the Medical Imaging and Computer Aided Diagnosis group of the Catholic University Nijmegen on mammographic image analysis (student exchanges).

Fig. 3: Left: Segmentation of an abdominal aortic aneurysm. Right: Calculation of the displacement (right, mm), VonMises strain (middle), Von Mises stress (kPa, right). [De Putter et al., 2005] Molecular Imaging of Ischemic Heart Disease In 2004 a major BSIK grant “Molecular Imaging of Ischemic Heart Disease” was rewarded to Univ. of Maastricht, TU/e, Philips and Organon. The BMIA group participates in the visualization of multi-spectal and tensor-based image data, classification of atherosclerotic plaque from multispectral MRI data, and the quantitative analysis of heart motion (multi-scale optic flow). Neurosurgical navigation In 2005 the Dept.of Neurosurgery at the Maastricht University Hospital acquired the first open low-field (0.15 Tesla) intra-operative MRI in the Netherlands. We collaborate with projects focused on the nonlinear warping of pre-operative MRI and CT data onto the intra-operative data, the integrated and interactive 3D visualization of DTI, fMRI and atlas data, and the optimization of image quality. These projects are carried out in collaboration with Medtronic Inc.

Fig. 4. So-called toppoints of an image, defined as singularity points in the multi-scale structure, can efficiently be exploited for image-based content retrieval, i.e. subscene retrieval from a complex scene. The interest points (white) of a query object (lower right) and a scene containing two rotated, scaled and occluded versions of the object, are succesfully found. Interest points that do not match are shown in grey. [Platel, Balmashnova et al., 2005]

The development of a Mathematica software library for rapid symbolic and numerical prototyping, ‘MathVisionTools’ A general Import format to deal with a wide spectrum of image formats, such as DICOM 3.0, FDF MRI data, 3D US data etc.; A symbolic and numerical generalized and speed-optimized toolbox for doing high order differential geometry on multi-dimensional images; A library of advanced image analysis functions.


64

Fig. 5. Illustrative Volume Rendering: Crosshatching using a smoothed direction field visualised with oriented "X" and "-" glyphs. This project focuses on applying techniques from illustrative rendering to (medical) volume datasets. In many cases, illustrative methods can create a clearer image than photorealistic methods, by leaving out unimportant details and emphasising important features using simple illustration primitives like points and lines [MSc project of Stef Busking, BME]. http://www.bmi2.bmt.tue.nl/image-analysis/Research/Projects/20010901-Visualization/index.html

External projects The Problem of Scale in Biomedical Image Analysis VICI grant of 1.25 M€ rewarded to dr. Luc Florack, dossier number NWO: 016.053.605, allocated on 01-01-2005. See the abstract in the description of the end of this report. Deep Structure and Singularities in Computer Vision. Starting date: 1 October 2002 Duration: 36 months Funding: European Community, 5th framework. Description: The group TUE-BMIA participates in an EU consortium (partners see below), whose overall objective is to develop sophisticated representations of images and shapes by merging principles and methods from 1) scale space theory, 2) singularity theory, and 3) algorithmics, and to create effective algorithms for solving computer vision tasks on the basis of these. URL: http://www.itu.dk/English/research/innovation/projects/DSSCV/ Budget total project: 1.5 M€. Partners: • IT University of Copenhagen (IT-C), Image Group, Department of Innovation • IT University of Copenhagen (IT-C), Algorithm Group, Theory Department • Technical University of Eindhoven (TU/e), BME - Biomedical Image Analysis • University of Liverpool (LU), Department of Mathematical Sciences • University of Copenhagen, 3D Lab (3D Lab), School of Dentistry Perceptual Grouping Period: Sept 2001 – Sep 2005 Funding: STW-SNN (1 PhD student) and TUE-BMT (1 PhD student) The objective of this project is to develop robust methods for segmenting contours in noisy images, inspired by the mechanisms recognized in human front-end vision. The focus is in particular on the development of new methods to understand the multi-scale nature of grouping perceptually equivalent elements, such as edge elements and textural features. Recently a new class of steerable wavelet orientation kernels is discovered (‘invertible orientation bundle’) with interesting properties for feature linking. Hemodyn Period: September 2002 – September 2006 Funding: Philips Medical Systems (1 PhD student) The prediction of the chance of rupture for the abdominal aorta is calculated by cardiovascular flow calculations. The data are specific for each patient, and derived from high resolution CT scans of the patient. Advanced methods in volume visualization Period: September 2002 – September 2006 Funding: Philips Medical Systems (2 PhD students) The high volumes of data both from volume-CT and new MRI 3D fast scanning protocols need for fast and efficient 3D volume presentations, as an effective form of data compression and communication with the referring physician and the surgeon. New algorithms are designed in the PMS Easy-Vision environment, particularly focusing on being able to cope with rapid answers and flexibility towards new questions of users.


Doctoral degrees 2005: Duits, R; Perceptual Organization in Image Analysis: a mathematical approach based on scale, orientation and curvature. TU Eindhoven, Eindhoven 2005, xiii+256 pp.ISBN 90-386-2747-5. Project funded by STW, project 4496 Papers in international journals M.A. van Almsick, R. Duits, E.M. Franken, B.M. ter Haar Romeny, From Stochastic Completion Fields to Tensor Voting , Lecture notes in computer science, 3753, 124-134, (2005) M.A. van Almsick, R. Duits, E.M. Franken, B.M. ter Haar Romeny, From Stochastic Completion Fields to Tensor Voting, Lecture notes in computer science, 3753, 124-134, (2005) E. Angelié, P.J.H. de Koning, M.G. Danilouchkine, H.C. van Assen, G. Koning, R.J. van der Geest, J.H.C. Reiber, Optimizing the automatic segmentation of the left ventricle in magnetic resonance images, Med. Phys., 32(2), 369-375, (2005) E. Balmachnova, L.M.J. Florack, B. Platel, F.M.W. Kanters, B.M. ter Haar Romeny, Stability of Top-Points in Scale Space, Lecture notes in computer science, 3459, 62-70, (2005) H. Bouma, A. Vilanova, L.J. van Vliet, F.A. Gerritsen, Correction for the Dislocation of Curved Surfaces caused by the PSF in 2D and 3D CT Images, PAMI, 27(9), 1501-1507, (2005) R. Duits, F.M.W. Kanters, L.M.J. Florack, B.M. ter Haar Romeny, A Comparison of the Deep Structure of ¨α-Scale Spaces, Lecture notes in computer science, 3753, 234-248, (2005) R. Duits, B.J. Janssen, F.M.W. Kanters, L.M.J. Florack, Linear Image Reconstruction from a sparse set of alpha scale space features by means of inner products of Sobolev type, Lecture notes in computer science, 3753, 96-111, (2005) R. Duits, M.A. van Almsick, M. Duits, R. Duits, E.M. Franken, Invertible Orientation Scores as an Application of Generalized Wavelet Theory, Lecture notes in computer science, 15(2), 101-140, (2005) M. Felsberg, R. Duits, L.M.J. Florack, The Monogenic Scale Space on a Rectangular Domain and its Features, IJCV, 64(2/3), 187-201, (2005) G.L.T.F. Hautvast, S. Lobregt, M. Breeuwer, A. Vilanova, F.A. Gerritsen, Automatic Contour Detection in Short-Axis Cardiac Cine MR Data , Journal of Cardiovascular Magnetic Resonance, 7(1), 323-324, (2005) A.M. Heemskerk, G.J. Strijkers (contact), A. Vilanova, M.R. Drost, K. Nicolay, Determination of mouse skeletal muscle architecture using three dimensional diffusion tensor imaging, Magn Reson Med, 53(6), 1333-1340, (2005) B.J. Janssen, F.M.W. Kanters, R. Duits, L.M.J. Florack, B.M. ter Haar Romeny, A Linear Image Reconstruction Framework Based on Sobolev Type Inner Products, Lecture notes in computer science, 3459, 85-95, (2005) F.M.W. Kanters, M. Lillholm, R. Duits, B.J.P. Jansen, B. Platel, L.M.J. Florack, B.M. ter Haar Romeny, On Image Reconstruction from Multiscale Top Points, Lecture notes in computer science, 3459, 431-439, (2005) B. Platel, M.. Fatih Demirci, A. Shokoufandeh, L.M.J. Florack, F.M.W. Kanters, B.M. ter Haar Romeny, S. Dickinson, Discrete Representation of Top Points via Scale Space Tessellation, Lecture notes in computer science, 3459, 73-80, (2005) B. Platel, E. Balmachnova, L.M.J. Florack, F.M.W. Kanters, B.M. ter Haar Romeny, Using Top-Points as Interest Points for Image Matching, Lecture notes in computer science, 3753, 211-222, (2005) S. de Putter, M. Breeuwer, U. Kose, et al., F.N. van de Vosse, F.A. Gerritsen, Automatic determination of the dynamic geometry of abdominal aortic aneurysm from MR with application to wall stress simulations, ICS, 1281, 339-344, (2005) Contributions to international conference proceedings H.C. van Assen, M.G. Danilouchkine, A.F. Frangi, S. Ordás, J.J.M. Westenberg, J.H.C. Reiber, B.P.F. Lelieveldt, SPASM: Segmentation of Sparse and Arbitrarily Oriented Cardiac MRI Data using a 3D-ASM, in FIMH 2005. In: Lecture Notes in Computer Science; Editors: A.F. Frangi, P. Radeva, A. Santos, and M. Hernandez, Barcelona, Spain, vol. 3504: 33-43, (2005) H.L.T. de Bliek, S. Lobregt, K. Visser, F.A. Gerritsen, Fast and Easy Whole-Heart MRA Segmentation and Visualization, in Proceedings of the 8th Annual Scientific Meeting of the SCMR, 21-23 January 2005, San Francisco, United States, 309-311, (2005)


66

H.L.T. de Bliek, S. Lobregt, K. Visser, F.A. Gerritsen, Fast and easy whole-heart MRA segmentation and visualization, in Proceedings CARS 2005, International Congress Series, Berlin, Germany, 1270, (2005) M.G. Danilouchkine, J.J.M. Westenberg, H.C. van Assen, J.H.C. Reiber, B.P.F. Lelieveldt, 3D Model-Based Approach to Lung Registration and Prediction of Respiratory Cardiac Motion, in MICCAI 2005. In: Lecture Notes in Computer Science; Editors: J. Duncan and G. Gerig, Palm Springs, United States, vol. 3750: 951-959, (2005) L.M.J. Florack, Deep Structure from a Geometric Point of View, in Deep Structure, Singularities, and Computer Vision: Vol. 3753, Lecture Notes in Computer Science; Editors: O. Fogh Olsen, L. M. J. Florack and A. Kuijper, Maastricht, Netherlands, 135-145, (2005) E.M. Franken, M.A. van Almsick, P.M.J. Rongen, L.M.J. Florack, B.M. ter Haar Romeny, Steerable Tensor Voting, in ASCI Conference 2005; Heijen, Netherlands, 65-72, (2005) B.M. ter Haar Romeny, Biomedical Image Analysis, in Proc. International Mathematica Symposium 2005; Editors: P. Abbott, Perth, Australia, 45-85, (2005) B.M. ter Haar Romeny, M.A. van Almsick, Edge Preserving Smoothing with Euclidean Shortening flow, in Proc.International Mathematica Symposium 2005; Editors: P. Abbott, Perth, Australia, 361-368, (2005) B.M. ter Haar Romeny, Dither Removal, in Proc. International Mathematica Symposium 2005; Editors: P.Abbott, Perth, Australia, 285-299, (2005) B.M. ter Haar Romeny, Biologically inspired multi-scale image analysis, in Proc. 4th Dutch Endo-Neuro-Psycho Meeting; Editors: H. Pijl, S. van Dam, B. Ellenbroek, M. van Turenhout, Doorwerth, Netherlands, 293, (2005) G.L.T.F. Hautvast, M. Breeuwer, S. Lobregt, A. Vilanova, F.A. Gerritsen, Automatic cardiac contour propagation in short axis cardiac MR images, in CARS; International Conference Series - Elsevier, Berlin, Germany, 351--356, (2005) M. Mlejnek, P. Ermes, A. Vilanova, R. van der Rijt, H. van den Bosch, F.A. Gerritsen, E. Gröller, Profile flags: A novel metaphor for probing of T2 maps, in IEEE Visualization; Minneapolis, United States, 599-606, (2005) B. Moberts, A. Vilanova, J.J. van Wijk, Evaluation of fiber clustering methods for diffusion tensor imaging, in IEEE Visualization; Minneapolis, United States, 65-72, (2005) S. Ordás, H.C. van Assen, J. Puente, B.P.F. Lelieveldt, A.F. Frangi, Parametric Optimization of a Model-Based Segmentation Algorithm for Cardiac MR Image Analysis: a Grid-Computing Approach, in From Grid to Healthgrid, Proceedings of Healthgrid 2005. Studies in health technology and informatics; Oxford, United Kingdom, vol 112: 146-156, (2005) S. Ordás, H.C. van Assen, L. Boisrobert, M. Laucelli, J. Puente, B.P.F. Lelieveldt, A.F. Frangi, Statistical Modeling and Segmentation in Cardiac MRI using a Grid Computing Approach, in European Grid Conference 2005. In: Lecture Notes in Computer Science; Editors: P.M.A. Sloot, A.G. Hoekstra, T. Priol, A. Reinefeld, and M. Brubak, Amsterdam, Netherlands, vol. 3470: 6-15, (2005) K.S.P. Steenstrup Pedersen, R. Duits, M. Nielsen, On Alpha Kernels, Levy Processes, and Natural Image Statistics, in Scale Space and PDE Methods in Computer Vision; Editors: Kimmel, Sochen, Weickert, Hofgeismar, Germany, 468-479, (2005) J.F. Peters, S.E. Grigorescu, R. Truyen, F.A. Gerritsen, A.H. de Vries , R.E. van Gelder, P. Rogella, Robust automasted polyp detection for low-dose and normal-dise virtual colonoscopy, in Proceedings CARS 2005, International Congress Series 1281, Berlin, Germany, 1146-1150, (2005) B. Platel, E. Balmachnova, L.M.J. Florack, F.M.W. Kanters, B.M. ter Haar Romeny, Using Top-Points as Interest Points For Image Matching, in Deep Structure, Singularities, and Computer Vision; Editors: O. Fogh Olsen, L. M. J. Florack and A. Kuijper, Maastricht, Netherlands, 211-222, (2005) Contribution to Books Fogh Olsen, L.M.J. Florack, A. Kuijper, Deep Structure, Singularities, and Computer Vision. First International Workshop, DSSCV 2005, Maastricht, The Netherlands, June 9-10 2005. In: Lecture Notes in Computer Science, vol. 3753., Springer-Verlag, ISBN 3-540-29836-3 (2005)


B.P.F. Lelieveldt, A.F. Frangi, S.C. Mitchell, H.C. van Assen, S. Ordás, J.H.C. Reiber, M. Sonka, 3D Active Shape and Appearance Models in Cardiac Image Analysis, in Mathematical Models in Computer Vision: The Handbook; Editors: N. Paragios, Y. Chen, O. Faugeras, 471-485, Springer, Book Chapter, ISBN 0387263713 (2005) A. Vilanova, S. Zhang, G. Kindlmann, D. Laidlaw, An Introduction to Visualization of Diffusion Tensor Imaging and its Applications, in Visualization and Image Processing of Tensor Fields, 121-153, Springer Verlag, Book Chapter, ISBN 3-540-25032-8 (2005) Technical reports: R. Duits, M.A. van Almsick, The Explicit Solutions of Linear Left Invariant 2nd order evolution equations on the 2D-Euclidean Motion Group, 1-37, CASA-report 05-43, Internal Report ISSN 0926-4507 (2005) E.M. Franken, Context-Enhanced Detection of Electrophysiology Catheters in Noisy Fluoroscopy Images, Internal BMIA Report (2005) Highlights The highlights in 2005 were the reward of the NWO VICI grant to dr. Luc Florack, the NWO VENI grant to dr. Anna Vilanova and the Cum Laude PhD of the first promovendus of our group, dr.ir. Remco Duits. VENI: dr.Luc Florack Title: The Problem of Scale in Biomedical Image Analysis. Summary:Improving capabilities and resolving power of medical imaging devices, the sheer number of acquisitions (a modern hospital produces tens of terabytes of image data per year), and their complex nature make a careful diagnosis by visual inspection cumbersome or even prohibitive. Growing expectations from society for better diagnoses (one third of a modern hospital's budget is dedicated to imaging) and the desire for elaborate precautionary screenings, e.g. for lung, prostate, breast and colon cancer, further aggravate the problem. Biomedical image analysis may come at rescue. Biomedical image analysis is intended to assist radiologists (CAD = Computer-Aided Diagnosis), surgeons (image guided intervention), and scientists in interpreting complex image data. The capability of a (trained) human expert to visually interpret the data (if suitably presented) is without parallel. Indeed, image analysis systems outperforming human visual expertise are exceptionally rare. The dif�culty in designing such systems lies in the fact that image analysis is a notoriously hard problem, for two reasons: 1. Specifc visual interpretation tasks are hard to translate into operational terms needed for algorithmic design. 2. Local image values depend on discretization and quantization details, but relevant information does not. The second problem is a generic one, common to all images regardless of task. It compels us to abstract from pixel details and, a fortiori, to introduce an a priori free scale (= inverse resolution) parameter to replace the irrelevant grid constant. Time-polished methodologies that account for scale explicitly are, among others, scale space theory, wavelet theory and mathematical morphology. In virtually all other generic paradigms scale enters in disguise, e.g. in statistical techniques. The generic problem can thus aptly be rephrased as the problem of scale. I intend to solve the problem of scale with a multidisciplinary team by

o unfication, extension, and application of scale paradigms, and o biomimicking, i.e. simulating human visual expertise.

I will provide proof of concept in the form of specfic multi-scale biomedical image analysis algorithms, inspired by models of the visual brain from neurophysiology and psychophysics. VENI: Dr. A. (Anna) Vilanova í Bartrolí Title: Visualization of global tensor information for diffusion tensor imaging Summary: During random diffusion driven displacements, molecules follow tissue structure at microscopic scale. Diffusion Tensor Imaging (DTI) is a recent magnetic resonance (MR) acquisition technique that measures water diffusion. This technique allows to visualize structures at a scale well beyond the usual 3D medical imaging scanner resolution. Diffusion is represented by a positive symmetric tensor of second order. Visualizing high-dimensional data is a challenging problem. Several techniques have been presented in the last years to visualize tensor fields. However, these techniques either require a simplification to a scalar or to a vector field; or they merely show local information (i.e.,at isolated points, but no neighborhood interactions). The goal of this proposal is research and development of visualization techniques for second-order tensor fields (specifically, for DTI) which reveal the complex spatial relationships of tensors (i.e., global information). The success of this research should provide and improve visualization and image processing techniques for tensor data. These techniques will ultimately help scientists and physicians to obtain a better understanding of the complex information of tensor fields. CUM LAUDE PhD defence: Remco Duits: Perceptual Organization in Image Analysis: a mathematical approach based on scale, orientation and curvature. TU Eindhoven, Eindhoven 2005, xiii+256 pp. ISBN 90-386-2747-5. Project funded by STW, project 4496. September 15, 2005.


68

2.3.6 Contribution of TUD-TNW-tn-qi Multidimensional Image Analysis We have continued our efforts in developing novel strategies for the processing, analysis and interpretation of multi-dimensional images, including time-series, color images and hyper-spectral images. The work focuses on the detection and subsequent characterization of geometric image structures rather than specific intensity patterns. This approach yields generic tools for image description that can be applied in a wide variety of applications. The methods are largely independent of the physical imaging technique used or the intrinsic image resolution. The theoretical component in this work has focused on · Interactive 3D segmentation using connected orthogonal contours · Model based segmentation of material mixtures near boundaries using an analytical scale-invariant model for the relation between intensity and gradient magnitude. · Model based segmentation of three-material mixtures near generalized T-junctions. · Level set PDE’s for modeling single-curved surfaces. · Segmentation and size measurements of bumps. · Improved wave-front modeling for fast marching. · Decomposition of the structure tensor near multi-model neighborhoods. · Quantitative evaluation of dynamic superresolution methods. · Characterization of pore spaces and network structures in tablets. · Motion and flow analysis of liquid flow in micro-machined nanoliter wells. · Design of orientation selective filters for lines, planes, edges in 3D space. · An addressable multi-resolution sampling grid for 3D orientation. · Dynamic super-resolution in image sequences based on adaptive (scale, orientation and curvature) data fusion of registered local image data. · Partial volume effects estimation in 3-D CT based on density and gradient trajectories. · Computer aided diagnosis by automatic screening for polyps in colonography. · Inter and intra patient registration of DTI images applied to atlas building, tractography and patient studies. · Using line segments as structuring elements for sampling-invariant measurements · The generalized Radon transform: sampling, accuracy and memory considerations The aforementioned developments have been applied in numerous industrial collaborations and medical applications. · Analysis of micro-structures formed by complex networks of bio-polymers; · Detection of channels and faults in 3D seismic images; · Analysis of liquid flow and diffision using tracer particles; · Analysis of perceptual image quality of printed materials using local image features, scale and color; · Analysis, transformation, visualization and registration of diagnostic and surgical images for use in minimally invasive surgery and placement of endoprostheses. · Virtual endoscopy; · Detection and progression analysis of glaucoma. · MRI thermography. · Diffusion Tensor MRI · Characterizing the Three- Dimensional Organization of Telomeres External projects 4D image processing for analyzing motion patterns of the wrist Period: 2005-2009; Funding STW In this project we will develop image processing and analysis methods for 4D characterization of the wrist. The extracted information will be combined with biomechanical models for diagnosis and treatment planning. Processing and analysis of Diffusion Tensor MRI at 3Tesla Period: 2005-2008; Funding VL-e through AMC Amsterdam In this project we will develop image processing and analysis methods for Diffusion Tensor MRI images. Dynamic superresolution of small moving objects Period: 2004-2007; Funding: TNO Investigating techniques for detection, registration and subsequent superresolution of small moving objects Intelligent image interpretation for automated quality grading of young tomato plants in horticulture Period: 2003-2006; Funding: IOP beeldverwerkingOntology driven image processing and analysis for the quality grading of young tomato plants in horticulture.


Detection and progression analysis of glaucoma Period: 2002-2005; Funding: LDT (Laser Diagnostic Technologies, recently taken over by Carl Zeiss Meditec). Early detection of glaucoma and its progression in images acquired through scanning laser polarimetry, by means of the GDx apparatus. Cyttron Period: 2003-2007; Funding: Bsik Nanophoresis Period: 2002-2010; Funding: FOM Superresolution in undersampled image sequences Period: 2002-2005; Funding: IOP beeldverwerking Increasing the spatial resolution in exchange for temporal resolution through adaptive techniques. Automatic polyp detection in 3D CT Period 2002-2006; Funding IOP beeldverwerking Image processing, analysis and recognition to facilitate automatic screening for polyps in 3D CT of the colon. Multi-dimensional measurement techniques Period: 2002-2006; Funding: FOM Fundamental techniques for multi-dimensional image processing and analysis. Improved techniques for virtual endoscopy Period: 2001-2005; Funding: Philips Medical Systems; Doctoral degrees Pekalska, E.; (cum laude) 17 January 2005; Dissimilarity representations in pattern recognition , TU Delft, Delft 2005, 322 pp., ISBN 90-9019021-X, External project (=2e geldstroom). Contributions to Books Vos, F.M., R.E. van Gelder, I.W.O. Serlie, J. Florie, Y.C. Nio, F.H. Post, R. Truyen, A.M. Vossepoel, and J. Stoker; Virtual Colonoscopy (Ch.9), in: J. Dankelman, C.A. Grimbergen, H.G. Stassen (eds.), Engineering for Patient Safety: Issues in Minimally Invasive Procedures, Lawrence Erlbaum Associates, Mahwah, NJ, 2005, 206-225. Papers in international journals Bouma, H., A. Vilanova, L.J. van Vliet, and F.A. Gerritsen; Correction for the dislocation of curved surfaces caused by the PSF in 2D and 3D CT images, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, no. 9, 2005, 1501-1507. Bruin, P.W. de, V.J. Dercksen, F.H. Post, A.M. Vossepoel, G.J. Streekstra, and F.M. Vos; Interactive 3D segmentation using connected orthogonal contours, Computers in Biology and Medicine, vol. 35, no. 4, 2005 (@2004), 329-346. Doel, L.R. van den, P.T. Nagy, L.J. van Vliet, and G.P. Neitzel; Regularized phase-tracker with iso-phase scanning strategy for analysis of dynamic interferograms of nonwetting droplets under excitation, Applied Optics (Optical Society of America), vol. 44, no. 14, 2005, 2695-2704. Fulwyler, M., Q.S. Hanley, C. Schnetter, I.T. Young, E.A. Jares-Erijman, D.J. Arndt-Jovin, and T.M. Jovin; Selective Photoreactions in a Programmable Array Microscope (PAM): Photoinitiated Polymerization, Photodecaging, and Photochromic Conversion, Cytometry, vol. 67A, no. 2, 2005, 68-75. Garini, Y., B.J. Vermolen, and I.T. Young; From micro to nano: recent advances in high-resolution microscopy, Current Opinion in Biotechnology, vol. 16, no. 1, 2005, 3-12. Garling, E.H., M. van Eck, T. Wedding, D.J. Veeger, E.R. Valstar, and R.G.H.H. Nelissen; Increased muscle activity to stabilise mobile bearing knees in patients with rheumatoid arthritis, The Knee, vol. 12, no. 3, 2005, 177-182. Garling, E.H., B.L. Kaptein, K. Geleijns, R.G.H.H. Nelissen, and E.R. Valstar; Marker Configuration Model-Based Roentgen Fluoroscopic Analysis, Journal of Biomechanics, vol. 38, no. 4, 2005, 893-901.


70

Kaptein, B.L., E.R. Valstar, B.C. Stoel, P.M. Rozing, and J.H.C. Reiber; A new type of model-based Roentgen stereophotogrammetric analysis for solving the occluded marker problem, Journal of Biomechanics, vol. 38, no. 11, 2005, 2330-2334. Kutchoukov, V.G., L. Pakula, G.O.F. Parikesit, Y. Garini, L.K. Nanver, and A. Bossche; Fabrication of nanofluidic devices in glass with polysilicon electrodes, Sensors and Actuators A, Physical, vol. 123-124 (Eurosensors XVIII 2004 - The 18th European Conf. on Solid-State Transducers), 2005, 602-607. Louis, S.F., B.J. Vermolen, Y. Garini, I.T. Young, A. Guffei, Z. Lichtensztejn, F. Kuttler, T.C.Y. Chuang, S. Moshir, V. Mougey, A.Y.C. Chuang, P.D. Kerr, T. Fest, P. Boukamp, and S. Mai; c-Myc induces chromosomal rearrangements through telomere and chromosome remodeling in the interphase nucleus, Proceedings of the National Academy of Sciences USA, vol. 102, no. 27, 2005, July 5, 9613-9618. Luengo Hendriks, C.L., M. van Ginkel, P.W. Verbeek, and L.J. van Vliet; The generalized Radon transform: sampling, accuracy and memory considerations, Pattern Recognition, vol. 38, no. 12, 2005, 2494-2505. Luengo Hendriks, C.L., and L.J. van Vliet; Using line segments as structuring elements for sampling-invariant measurements, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, no. 11, 2005, 1826-1831. Mai, S., and Y. Garini; Oncogenic Remodeling of the Three-Dimensional Organization of the Interphase Nucleus: c-Myc Induces Telomeric Aggregates Whose Formation Precedes Chromosomal Rearrangements, Cell Cycle, vol. 4, no. 10, 2005, 1327-1331. Melissant, C., A. Ypma, E.E.E. Frietman, and C.J. Stam; A method for detection of Alzheimers disease using ICA-enhanced EEG measurements, Artificial Intelligence in Medicine, vol. 33, no. 3, 2005, 209-222. Minekus, J.P.J., E.R. Valstar, P.M. Rozing, M.J. de Vos, R.L. Diercks, W.J. Willems, and J. Dankelman; Factors influencing the surgical process during shoulder joint replacement: Time-action analysis of five different prostheses and three different approaches, Medical Science Monitor: Clinical Research, vol. 11, no. 1, 2005, CR14-20. Moerman, R., J. Knoll, C. Apetrei, L.R. van den Doel, and G.W.K. van Dedem; Quantitative Analysis in Nanoliter Wells by Prefilling of Wells Using Electrospray Deposition Followed by Sample Introduction with a Coverslip Method, Analytical Chemistry, vol. 77, no. 1, 2005, 225-231. Parikesit, G.O.F., A.P. Markesteijn, V.G. Kutchoukov, O. Piciu, A. Bossche, J. Westerweel, Y. Garini, and I.T. Young; Electroosmotic flow analysis of a branched U-turn nanofluidic device, Lab on a Chip, vol. 5, no. 10, 2005, 1067-1074. Valstar, E.R., R. Gill, L. Ryd, G. Flivik, N. Borlin, and J. Karrholm; Guidelines for standardization of radiostereometry (RSA) of implants, Acta Orthopaedica, vol. 76, no. 4, 2005, p. 563-572. Vermeer, K.A., N.J. Reus, F.M. Vos, A.M. Vossepoel, and H.G. Lemij; Automated detection of wedge-shaped defects in polarimetric images of the retinal nerve fiber layer, Eye, vol. 19, 2005, 1-9 (at the press). Vermolen, B.J., Y. Garini, S. Mai, V. Mougey, T. Fest, T.C.-Y. Chuang, A.Y.-C. Chuang, L. Wark, and I.T. Young; Characterizing the Three- Dimensional Organization of Telomeres, Cytometry, vol. 67A, no. 2, 2005, 144-150. Wu, Y.S., H.W. Frijlink, L.J. van Vliet, L. Stokroos, and K. van der Voort Maarschalk; Localization dependent analysis of porosity and pore direction in tablets, Pharmaceutical Research, vol. 22, no. 8, 2005, 1399-1405. Contributions to international conference proceedings Blaas, J., C.P. Botha, B. Peters, F.M. Vos, and F.H. Post; Fast and Reproducible Fiber Bundle Selection in DTI Visualization, IEEE Visualization 2005 (Proc. Conf. Mineapolis, MN, USA, Oct.23-25), ACM Press, 2005, p. accepted. Dietrich, H.R.C., L.R. v.d. Doel, W. Hoyer, W. van Oel, G. Liqui Lung, Y. Garini, T. Jovin, and I.T. Young; Adaptation of Nanoarrays for the Study of a-synuclein aggregation - preliminary results, in: D.V. Nicolau, J. Enderlein, R.C. Leif, D.L. Farkas, R.Raghavachari (eds.), Imaging, Manipulation, and Analysis of Biomolecules and Cells: Fundamentals and Applications III (Proc. SPIE Photonics West: Biomedical Optics (BiOS) San Jose, California, USA, Jan.22-27), Proc. SPIE, vol. 5699, 2005, 395-402. Dietrich, H.R.C., E. van IJsseldijk, S. de Vries, I.T. Young, and Y. Garini; Advances in the Development of a Novel Method to be used in Proteomics using Gold Nanoparticles, Proc. Int. Conf. on Biopartitioning and Purification (SL, June 20-24), 2005, 17-18


Dietrich, H.R.C., I.T. Young, and Y. Garini; Gold nanoparticles: A novel application of spectral imaging in proteomics - preliminary results, in: G.H. Bearman, Anita Mahadevan-Jansen, Richard M. Levenson (eds.), Spectral Imaging: Instrumentation, Applications, and Analysis III (Proc. SPIE Photonics West: Biomedical Optics (BiOS) San Jose, California, USA, Jan.22-27), Proc SPIE, vol. 5694, 2005, 82-89. Dijkers, J.J., C. van Wijk, F.M. Vos, J. Florie, Y.C. Nio, H.W. Venema, R. Truyen, and L.J. van Vliet; Segmentation and size measurement of polyps in CT colonography, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 (Proc. 8th Int. Conf., Palm Springs, CA, USA, Oct. 26-29), Lecture Notes in Computer Science, vol. 3749, Springer Verlag, Heidelberg, 2005, 712-719. Docter, M.W., I.T. Young, V.G. Kutchoukov, A. Bossche, P.F.A. Alkemade, and Y. Garini; A novel concept for a mid-field microscope, in: Tuan Vo-Dinh, J.R. Lakowicz, Z.K. Gryczynski (eds.), Plasmonics in Biology and Medicine II (Proc. SPIE Photonics West: Biomedical Optics (BiOS) San Jose, California, USA, Jan.22-27), Proc. SPIE, vol. 5703, 2005, 118-126. Parikesit, G.O.F., V.G. Kutchoukov, A. Bossche, I.T. Young, and Y. Garini; Optical detection of electrokinetically manipulated single molecules in a nanofluidic chip, in: I. Papautsky, I. Chartier (eds.), Proc. Conf. Microfluidics, BioMEMS, and Medical Microsystems III, Proc. SPIE, vol. 5718, 2005, 133-141. Pham, T.Q., M. Bezuijen, L.J. van Vliet, K. Schutte, and C.L. Luengo Hendriks; Performance of optimal registration estimators, in: Z. Rahman, R.A. Schowengerdt, S.E. Reichenbach (eds.), Visual Information Processing XIV (Proc. SPIE Defense and Security Symposium, Orlando, Florida, USA, Mar.29-Apr.1), Proc. SPIE, vol. 5817, 2005, 133-144. Pham, T.Q., and L.J. van Vliet; Separable bilinear filtering for fast video preprocessing, CD Proc. IEEE Int. Conf. on Multimedia & Expo (Amsterdam, July 6-8), 2005, 1-4. Pham, T.Q., L.J. van Vliet, and K. Schutte; Influence of signal-to-noise ratio and point spread function on limits of super-resolution, in: E.R. Dougherty, J.T. Astola, K.O. Egiazarian (eds.), Image Processing: Algorithms and Systems IV (Proc. IS&T/SPIE's 17th Annual Symposium Electronic Imaging, San Jose, California, USA, Jan.1620), 2005, 169-180. Pham, T.Q., and L.J. van Vliet, Blocking artifacts removal by a hybrid filter method, in: B.J.A. Krose, H.J. Bos, E.A. Hendriks, J.W.J. Heijnsdijk (eds.), ASCI 2005; Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging (Heijen, NL, June 8-10), ASCI, Delft, 2005, 372-377. Rollano-Hijarrubia, E., F. van der Meer, A. van der Lugt, H. Weinans, H.A. Vrooman, A.M. Vossepoel, and R. Stokking; Improving the imaging of calcifications in CT by histogram-based selective deblurring, in: M.J. Flynn (eds.), Physics of Medical Imaging (Proc. Conf. San Diego, CA, USA, Feb.13-15), Proc. SPIE, vol. 5745-09, 2005, 67-78. Suprijanto, M.W. Vogel, F.M. Vos, H.A. Vrooman, and A.M. Vossepoel; Inter-frame Motion Correction for MR Thermometry, in: J. Duncan, G. Gerig (eds.), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 (Proc. 8th Int. Conf., Palm Springs, CA, USA, Oct. 26-29), Lecture Notes in Computer Science, vol. 3749, Springer Verlag, Heidelberg, 2005, 580-588. Vermolen, B.J., I.T. Young, A. Chuang, L. Wark, T. Chuang, S. Mai, and Y. Garini; Three-dimensional analysis tool for segmenting and measuring the structure of telomeres in mammalian nuclei, in: D.V. Nicolau, J. Enderlein, R.C. Leif, D.L. Farkas, R.Raghavachari (eds.), Imaging, Manipulation, and Analysis of Biomolecules and Cells: Fundamentals and Applications III (Proc. SPIE Photonics West: Biomedical Optics (BiOS) San Jose, California, USA, Jan.22-27), Proc. SPIE, vol. 5699, 2005, 111-120. Cooperations within ASCI With : TUD-EWI-mm-cgcc, TUD-EWI-mm-ict, UL-LUMC-lkeb and with EU-EMCR-bigr HIGHLIGHT Our publication in PNAS in 2005 (see below) represented a clear illustration of the power of applying quantitative, multi-dimensional image analysis tools that we developed to the study of cell replication in normal and malignant cell lines. Having participated in the discovery of the “telomeric disc” (published in BMC Biology in 2004), we showed in this work how this disc’s development is altered by the protein c-Myc and how this is related to chromosomal rearrangement. Relevant publication: Louis, S.F., B.J. Vermolen, Y. Garini, I.T. Young, A. Guffei, Z. Lichtensztejn, F. Kuttler, T.C.Y. Chuang, S. Moshir, V. Mougey, A.Y.C. Chuang, P.D. Kerr, T. Fest, P. Boukamp, and S. Mai; c-Myc induces chromosomal rearrangements through telomere and chromosome remodeling in the interphase nucleus, Proceedings of the National Academy of Sciences USA, vol. 102, no. 27, 2005, July 5, 9613-9618.


72

2.3.7 Contribution of UL-LIACS Cyttron Bio-Computing Search Project The Cyttron (www.cyttron.nl) consortium wants to implement a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail. We would like to provide a generic tool for identifying the molecular causes of disease, essential for the prevention of disease and the development of new drug and therapies, and to establish a platform for advanced diagnosis and tuning of individualized therapy, increasing effectiveness in health care. The consortium is highly multidisciplinary including (bio-)physicists, chemists, mathematicians, bio-informatics and image processing specialists, cell biologists, microscopists and medical researches from various research institutes. This sub-project of Cyttron focuses on image search algorithms and methods for bio-image data bases. External Projects Cyttron Bio-Computing Search Project 2004-2008, Bsik, 8.8 MEuro. The Cyttron consortium wants to implement a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail. We would like to provide a generic tool for identifying the molecular causes of disease, essential for the prevention of disease and the development of new drug and therapies, and to establish a platform for advanced diagnosis and tuning of individualized therapy, increasing effectiveness in health care. Contributions to international conference proceedings N. Sebe, E.M. Bakker, I. Cohen, T. Gevers, and T. Huang, Bimodal Emotion Recognition, proceedings 5th International Conference on Methods and Techniques in Behavioral Research, Measuring Behaviour 2005.

2.3.8 Contribution of UvA-FdNWI-ias Perception for autonomous systems We develop methodologies for accurate motion estimation and interpretation from image sequences. Mobile vision platforms have their applications in traffic and driving in unstructured terrain. Applications of static platforms are in public safety and intelligent care homes for the elderly. We cooperate with TNO Defense, Security and Safety on a number of projects, which are related to autonomous systems, such as the RoboJeep autonomous robot vehicle. We have continued our investigation into terrain classification. This is an important topic for off-road autonomous robot vehicle guidance. Range based sensor systems, such as stereo vision, cannot distinguish between solid obstacles such as rocks or soft obstacles such as tall patches of grass. Terrain classification is needed to prevent that the robot vehicle is stopped needlessly by the obstacle detection system. It can also be used to recognize sand roads or other drivable areas. Possible image features that can be exploited are color and texture. Using color recognition outdoors is difficult, because the observed color of a material is heavily influenced by environment conditions such as the scene composition and illumination. Previously, we have developed an approach that distinguished between different environment states in order to reduce the variation in to be recognized material colors. This approach has been extended so that additional features such as texture can be exploited. Further improvements have also eliminated the need for setting thresholds during the environment clustering stage. Also in cooperation with TNO D&V methodologies for public safety are developed. Many surveillance algorithms consist of different parts, such as object detection, segmentation, tracking, and recognition. Literature focuses on algorithms for one of these individual tasks. Integration of and communication between these tasks is often rather ad-hoc. In this project we worked on a statistical framework for visual tracking applications. It allows easy adaptation of the tracking application by substitution of one of the algorithms for part of the problem, without altering the remainder of the application. Furthermore, the framework uses minimum cost classification and feedback for updating the models using knowledge available elsewhere in the application. A common problem for surveillance algorithms is the occurrence of shadows. A novel technique for shadow detection has been developed. It is based on the assumption that often scenes are illuminated by several light sources, each with a different color. This gives rise to a color shift for shadow, as not all light sources are occluded to the same amount. The developed technique estimates this color shift and allows more accurate distinction between shadow and foreground objects. Visual perception of humans and their activities The ability to recognize humans and their activities by vision is key for a machine to interact intelligently and effortlessly with a human-inhabited environment. There are numerous important applications ranging from public safety, elderly care and intelligent vehicles to human motion capture/analysis. Two projects have started this year in the area of Looking-at-People and public safety. In the MultimediaN – Safety Pilot project (07/05-06/09) the initial focus is on 3-D human pose estimation from single images. We are looking into approaches that use a template hierarchy to match hundred of thousands of possible poses in near real-time. In particular, we are investigating appropriate matching metrics for hierarchical detection and subsequent verification stage, and ways to incorporate prior scene knowledge.


The CASSANDRA (Context-Aware SenSing for AggressioN Detection and Risk Assessment, 09/05-08/09) project, funded by NWO and in collaboration with the University of Groningen, aims to detect aggression in dynamic environments. Because events associated with the build-up or enactment of aggression are difficult to detect by a single sensor modality (e.g. shouting versus hitting someone) we investigate the fusion of video- and audio- sensing. Learning, probabilistic, and neural computing We develop learning methods and probabilistic reasoning methods for intelligent systems operating in a real world. In 2005 there was an increased focus on intelligent environments, which observe the persons moving in it, and have a ‘context awareness’. We focussed on two applications areas. The first one is a surveillance application with many cameras that do not have overlapping field of views. We developed a person tracking method that outperforms conventional (multi-hypothesis) tracking methods. A patent on the method is pending and there is strong interest in licenses. The STW project ended in 2005, but a proposal for an STW ‘Valorization grant’ was honoured (BaViS: Bayesian Visual Surveillance). The second application area is ‘cognitive robotics.’ Robots tend to move out of the factory floor into our daily lives. The IAS group is one of the 8 partners in a 9 MEuro European Integrated Project in which a ‘robot servant’ is developed. Novel methods for spatial cognition and space learning have been developed and are implemented. Another line of research concerns sensor data fusion. Typical real-world data have several modalities. In the past we have developed powerful probabilistic methods for sensor fusion in robotics or tracking. In a newly started project, funded by MultimediaN, we develop methods for fusion of audio and video data in multimodal information streams (such as movies, webcams, surveillance cameras) in order to create a robust tool for the detection of scene transitions and events, for example to detect, identify and track humans, or to find scenes with a predefined situational signature such as aggressive scenes, romantic scenes, etc. Principles of autonomous systems In the project on geometric algebra, we have continued the development of the conformal model as a compact language for programming geometry. We have deepened the work on efficient implementation, and on facilities for visualization of and experimentation with geometric algebra and geometry. This involves compiler and interpreter construction, and is a cooperation with the Faculty of Computer Science at the Free University (VU) Amsterdam. We currently have the world's fastest implementation of geometric algebra (by an order of magnitude). A simple ray tracer is about 20% slower than the optimal classical implementation of the same algorithms, while having much more structured code. Considerable effort was spent on finishing a tutorial book for the computer science audience, to help spread the distribution of the techniques to the field. Decision making in single- and multi-agent systems In the context of agent decision making under uncertainty, we have developed Perseus, the world’s fastest approximate POMDP solver. We recently extended Perseus to planning domains in which an agent has a continuous set of actions at its disposal, and to domains involving continuous states, which is relevant for planning in robotic domains. We also explored methods for decentralized planning under uncertainty for teams of communicating agents. Here we studied scenarios in which the agents have the ability to communicate, but bandwidth is limited and communication has a certain cost. Our work has also focused on algorithms for large-scale multi-agent coordination. Here we use the framework of coordination graphs which allows for a tractable approach to the coordination problem, by decomposing the global payoff function of the system into a sum of local terms. Recent work here involves (i) message passing techniques for approximate decision making (similar to belief propagation in Bayesian networks), (ii) distributed cooperative reinforcement learning (Q-learning) in large agent networks by decomposing the action-value using a coordination graph, and (iii) automatically learning the dependencies between the agents in cooperative multi-agent systems. Parts of this framework were incorporated in our UvA Trilearn RoboCup simulation team which reached 10th place at the 2005 World Championship in Osaka, Japan. The accompanied paper at the RoboCup Symposium describing our message passing techniques won the best scientific paper award. We have applied our experience to the soccer game when performed by a team of 4-Legged robots. The 4-Legged Soccer League is an ideal challenge, because it defines a standard hardware platform, which allows benchmark algorithms to perceive the environment, to reason about the optimal (joint) actions, and to perform those actions in a smooth way. This year we had to concentrate our effort to the localization algorithm, to cope with a field with no boundaries. The vertical structures on the field are two brightly colored goals and four uniquely colored flags that can be used by the robots for a coarse localization. When making use of the white field lines on the carpet, it has been shown that the error in localization can be pushed below six centimeters. Our group also contributed to the Variable Lighting Challenge, by using algorithm multiple pre-calibrated color-cluster lookup tables for different lighting conditions and switched between them dynamically (based on a brightness estimate). This approach is illustrated in the figure. The soccer world is dynamic, but has a fixed number of agents, and small playing field. To study the influence of larger real-world models, we focused on cooperation between teams of rescue-agents in a city after a disaster. As part of this research,


74

a preliminary literature study is made with respect to models and algorithms for multi-agent planning under partial observability. The focus has been on decentralized partially observable Markov decision processes (Dec-POMDPs), but also the more general case is studied in which agent's have their own pay-off function (Partially observable stochastic games - POSGs). The benefit of both approaches is their clean way to represent a multi-agent decision problem. The down-side is that solving them (i.e. finding the optimal joint policy) is provably intractable (NEXP-complete). Therefore, exact algorithms, of which only few exist, are limited to extremely small problems. To tackle realistic problems, approximation models and algorithms have to be developed. Another research line involves `Interactive Hierarchical Awareness’ in a multi-agent system. Modeling and control of such a system may benefit immensely from representing it at multiple levels of abstraction, i.e., using a hierarchy of representations. This means that the system's state is represented at multiple resolutions (in terms of time and abstraction) simultaneously for different levels in the hierarchy, and similarly, that actions are taken at multiple resolutions. In the Interactive Collaborative Information Systems (ICIS) project we aim at developing scalable and theoretically sound methods for constructing and updating hierarchical state representations for multi-agent systems. We are working on establishing criteria for hierarchical state representations, for instance based on action utilities, the Markov property, uncertainty, etc. Our main application domain is road traffic management. Distributed perception and sensor fusion Intelligent process control and decision making in complex systems require adequate situation assessment, which in turn requires processing of large amounts of heterogeneous information originating from different, spatially dispersed sources, such as, sensory systems, human observers, databases, etc. However, such "sensor fusion" is not trivial, since it requires adequate mapping between very heterogeneous concepts, we are confronted with noisy information sources and, due to large amounts of information, significant processing resources might be required (i.e. computational bottlenecks). Another characteristic of the domains we are focusing on is that constellations of information sources can change frequently and, prior to the operation, we never know which information sources will be available. In addition, such fusion systems often provide results which have a critical impact on the decision making process and, consequently, further course of events. Therefore, high quality of fusion results and prevention of misleading results is indispensable. In order to be able to deal with the mentioned challenges, we have introduced Distributed Perception Networks (DPN), multi-agent systems which support fusion based on distributed Bayesian Networks (BN). This work was carried out within the project Combined Systems at the Decis research institute. In this context, our research is focused primarily on the following problems: (i) Task driven self-configuration of DPN agents at runtime, which allows adaptation to dynamic information source constellations and supports reuse of partial fusion results. (ii) Efficient and robust information fusion with distributed BNs, which provide adequate mapping between observable events and beliefs in hypotheses about hidden events. (iii) Resource allocation in distributed fusion systems based on information theoretic criteria. (iv) Approaches to improved fusion accuracy, such as fail-safe design of fusion systems as well as localization of faulty model components and information sources. We find applications of our research in the field of crisis management. Decentralized information sources can fuse information and extract the relevant events in the environment. Such a set-up is depicted in the figure. External Projects Geometric Algebra: a New Foundation for Geometric Programming 2000-2006, 637kf This projects makes geometric algebra into a practical tool for geometric programming applications within computer science, notably robotics, vision and computer graphics. New techniques for specification of geometric programs are being developed, and a freely available fast implementation constructed. This is an NWO-sponsored project. Probabilistic models for distributed surveillance 2001-2005, STW, 545kf This project studies how probabilistic models like belief networks can be used in distributed surveillance system, where moving objects have to be tracked by a system of cameras with non-overlapping views.The information from the different camera systems will have to be combined. A number of fundamental problems has to be solved such as the identification problem (`is the object which is observed by one camera the same as the object observed by some other camera some time ago?') and the representation problem (`how can probabilistic information be integrated in a distributed system?'). Distributed User Modeling for Personalized Exploring Recommender Systems (DUMPERS) 2002-2006, NWO(ToKen2000), 225kE In the project DUMPERS we will develop methods for adding adaptive navigation support for users of web sites. Goal of this project is to extend existing methods with extra components like exploration of the effect of recommendation by using techniques from the field of reinforcement learning. We will use locally stored user models to provide the information about the individual users needed for personal recommendation without violating the privacy of the users.


COMBINED systems 2002-2006, 800kE In this project we study innovative methods for disaster management. In particular aspects concerning distributed observation systems are investigated. An important property of distributed observation systems is that they autonomously extract information from the monitored area. This information is shared with other services in the disaster management platform. The project is conducted in cooperation with TUDelft, Thales en TNO within the DECIS laboratory. Environment Representation for Autonomous Vehicles 2001-2005, 200kf In this project we develop a "RoboJeep" robotic vehicle. This vehicle must be able to navigate and perform useful observations autonomously in an unstructured environment. Different sensors such as stereovision cameras, ultra-sonic sensors, odometers, and a laser range scanner are applied for this goal. The objective of this research is to develop techniques for acquiring reliable representations of the environment from these sensors. Collaboration with TNO. A structure for maintaining a shared world model in a dynamic environment between differentiated embedded systems and their interaction with human supervisors 2001-2006, 490 kf We study embedded autonomous systems in distributed environments, for applications in public safety; monitoring and control of traffic and environmental conditions; assistance and clean-up work in disaster areas. Collaboration between systems requires a collective world model, and we develop methods for its consistency maintenance, in time-critical situations. We use robot soccer (RoboCup) as a case study. This project is a collaboration with the VU and is sponsored by Progress. COGNIRON: Cognitive Robot Companion 2004-2008, 600kE In this 6th framework project within IST we develop a cognitive robot which interacts with humans and has conceptual representations. Our group works on human activity analysis and on cognitive representations of objects and space. Interactive Collaborative Information Systems 004-2009, 736 k€ The ICIS project is an attempt to bridge the gap between traditional information capture, transmission and transaction processing to full-blown intelligent information systems techniques. This is a road starting with traditional information management techniques such as better structuring and managing of information (including data mining and data warehousing) and the development of self-generating software through to artificial intelligence and agent based systems and complete integration of technologies with human and social systems. Contributions to Books W. Zajdel, N. Vlassis, and B. J. A Kröse. Bayesian methods for tracking and localization. In E. Aarts, J. Korts, and W. Verhaegh, editors, Intelligent Algorithms, pages 243-258. Kluwer Academic Publishers, 2005. Papers in international journals B. Bakker.The concept of circular causality should be discarded. Commentary on Marc D. Lewis: Bridging emotion theory and neurobiology through dynamic system modeling. Behavioral and Brain Sciences, 28:195-196, 2005. O. Booij and H. T. Nguyen.A gradient descent rule for spiking neurons emitting multiple spikes. Information Processing Letters, 95(6):552-558, September 2005. L. Dorst Error propagation of the Procrustes method for 3-D attitude estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(2):221-229, February 2005. Jelle R. Kok, Matthijs T. J. Spaan, and Nikos Vlassis.Non-communicative multi-robot coordination in dynamic environments. Robotics and Autonomous Systems, 50(2-3):99-114, February 2005. J.M. Porta and E. Celaya. Reinforcement learning for agents with many sensors and actuators acting in categorizable environments.Journal of Artificial Intelligence Research, 23:79-122, 2005. J.M. Porta and B. J. A Kröse. Appearance-based concurrent map building and localization. Robotics and Autonomous Systems, 54(2):159-164, 2005. J.M. Porta, J.J. Verbeek, and B.J.A. Kröse. Active appearance-based robot localization using stereo vision. Autonomous Robots, 18(1):59-80, 2005. Matthijs T. J. Spaan and Nikos Vlassis. Perseus: Randomized point-based value iteration for POMDPs. Journal of Artificial Intelligence Research, 24:195-220, 2005. J. J. Verbeek, N. Vlassis, and B. J. A. Kröse. Self-organizing mixture models. Neurocomputing, 63:99-123, 2005.


76

W. Zajdel and B. J. A. Kröse. A sequential Bayesian algorithm for surveillance with non-overlapping cameras. Int. Journal of Pattern Recognition and Artificial Intelligence, 19(8):977-996, 2005. Contributions to international conference proceedings B. Bakker, M. Steingrover, R. Schouten, E. Nijhuis, and L. Kester. Cooperative multi-agent reinforcement learning of traffic lights. In Proc. of the Workshop on Cooperative Multi-Agent Learning, European Conference on Machine Learning, pages 24-36, 2005. B. Bakker, Z. Zivkovic, and B.J.A. Kröse. Hierarchical dynamic programming for robot path planning. In Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 3720-3725, 2005. O. Booij, Z. Zivkovic, and B. Kröse. Pruning the image set for appearance based robot localization. In Proceedings of the Annual Conference of the Advanced School for Computing and Imaging, pages 57-64, June 2005. A. T. Cemgil, W. Zajdel, and B. Kröse. A hybrid graphical model for robust feature extraction from video. In C. Schmid, S. Soatto, and C. Tomasi, editors, IEEE Computer Vision and Pattern Recognition (CVPR), pages 1158-1165, San Diego, June 2005. P. de Oude, J. Nunnink, and G. Pavlin. Distributed bayesian networks in highly dynamic agent organizations. In Proc. Belgian Dutch AI Conf., pages 104-111, Brussels, Belgium, 2005. P. Jansen, W. van der Mark, J.C. van der Heuvel, and F.C.A. Groen. Colour based off-road environment and terrain type classification. In Proc. IEEE ITSC 8th Int. Conf. on Intelligent Transport Systems, pages 61-66, Vienna, Austria, September 2005. G. Klaassen, W. Zajdel, and B.J.A. Kröse. Speech-based localization of multiple persons for an interface robot. In Proc. of IEEE Int. Conference on Computational Intelligence in Robotics and Automation (CIRA2005), pages 47-52, 2005. Jelle R. Kok, Pieter Jan 't Hoen, Bram Bakker, and Nikos Vlassis. Utile coordination: Learning interdependencies among cooperative agents. In Proc. of the IEEE Symposium on Computational Intelligence and Games (CIG'05), pages 29-36, Colchester, United Kingdom, April 2005. Jelle R. Kok and Nikos Vlassis. Using the max-plus algorithm for multiagent decision making in coordination graphs. In RoboCup 2005: Robot Soccer World Cup IX, Osaka, Japan, July 2005. Best Scientific Paper Award. Jelle R. Kok and Nikos Vlassis. Using the max-plus algorithm for multiagent decision making in coordination graphs: Extended abstract. In K. Verbeeck, K. Tuyls, A. Now'e, B. Manderick, and B. Kuijpers, editors, Proc. of the 17th Belgian-Dutch Conference on Artifical Intelligence, pages 359-360, October 2005. W. Kowalczyk and N. Vlassis. Newscast EM. In L. K. Saul, Y. Weiss, and L. Bottou, editors, Advances in Neural Information Processing Systems 17, pages 713-720. MIT Press, Cambridge, MA, 2005. J. Nunnink and G. Pavlin. Accuracy of sequential bayesian information fusion. In Proc. IASTED Int. Conf. on Artificial Intelligence and Applications, pages 545-550, Innsbruck, Austria, 2005. J. Nunnink and G. Pavlin. A probabilistic approach to resource allocation in distributed fusion systems. In Proc. Int. Conf. on Autonomous Agents and Multi-Agent Systems, pages 846-852, Utrecht, The Netherlands, 2005. Frans Oliehoek, Matthijs T. J. Spaan, and Nikos Vlassis. Best-response play in partially observable card games. In Benelearn 2005: Proceedings of the 14th Annual Machine Learning Conference of Belgium and the Netherlands, pages 45-50, February 2005. Frans Oliehoek, Nikos Vlassis, and Edwin de Jong. Coevolutionary Nash in poker games. In Proceedings of the 17th Belgian-Dutch Conference on Artifical Intelligence, pages 188-193, Brussels, Belgium, October 2005. G. Pavlin, P. de Oude, and J. Nunnink. A MAS approach to fusion of heterogeneous information. In Proc. Int. Conf. on Web Intelligence, pages 802-804, 2005. P.J.Withagen, F.C.A.Groen, and K.Schutte. Ccd characterization for a range of color camera's. In Proceedings Instrumentation and Measurement Conference, IMTC 2005, pages 2232-2235, Ottawa, Ontario, Canada, May 2005.


Josep M. Porta, Matthijs T. J. Spaan, and Nikos Vlassis. Robot planning in partially observable continuous domains. In K. Verbeeck, K. Tuyls, A. Now'e, B. Manderick, and B. Kuijpers, editors, Proc. of the 17th Belgian-Dutch Conference on Artifical Intelligence, pages 375-376, Brussels, Belgium, October 2005. Josep M. Porta, Matthijs T. J. Spaan, and Nikos Vlassis. Robot planning in partially observable continuous domains. In Proc. Robotics: Science and Systems, pages 217-224. MIT Press, 2005. B. Slamet and A. Visser. Purposeful perception by attention-steered robots. In Proc. 17th Dutch-Belgian Artificial Intelligence Conference, BNAIC'05, pages 209-215, Brussels, October 2005. Matthijs T. J. Spaan and Nikos Vlassis. Planning with continuous actions in partially observable environments. In Proceedings of the IEEE International Conference on Robotics and Automation, pages 3469-3474, Barcelona, Spain, April 2005. M. Steingrover, R. Schouten, S. Peelen, E. Nijhuis, and B. Bakker. Reinforcement learning of traffic light controllers adapting to traffic congestion. In Proc. of the Belgium-Netherlands Artificial Intelligence Conference, BNAIC'05, pages 216-223, 2005. M. van Someren and S. ten Hagen. Machine learning and reinforcement learning. In Bogdan Gabrys, Kauko Leiviska, and Jens Strackeljan, editors, Do Smart Adaptive Systems Exist?, volume 173 of Studies in Fuzziness and Soft Computing, pages 81-104. Springer-Verlag, January 2005. N. Vlassis, Y. Sfakianakis, and W. Kowalczyk. Gossip-based greedy Gaussian mixture learning. In Lecture Notes in Computer Science, volume 3746, pages 349-359. Springer-Verlag, 2005. W. Zajdel, Z. Zivkovic, and B.J.A. Kröse. Keeping track of humans: have I seen this person before? In Proc. Int. Conference on Robotics and Automation (ICRA), pages 2093-2098, 2005. Z. Zivkovic, B. Bakker, and B.J.A. Kröse. Hierarchical map building using visual landmarks and geometric constraints. In Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 7-12, 2005. Z. Zivkovic and B.J.A. Kröse. On matching interest regions using local descriptors - can an information theoretic approach help? In Proc. British Machine Vision Conference, pages 50-58, 2005. Experimental software The Dutch AIBO Team - Osaka 2005 Source Code, C++, 2005 http://www.dutchaiboteam.nl/robocup/robocup2005/DT2005-diff-GT2004.zip Source code for the Dutch robot dog team participating in the world championships. Patent W. Zajdel, A.T. Cemgil, B.J.A. Kröse A hybrid graphical model for on-line multicamera tracking, Europese octrooiaanvrage 04077917.5 Cooperation within ASCI RoboCup/Progress: with TUD -TN en VU, also UU (Siks) and TUD- ET (DISC).

2.3.9 Contribution of EUR-RMI-bigr Cardiovascular Image Analysis Vascular imaging has gone beyond the traditional depiction of vascular luminal morphology. State-of-the art imaging techniques have the potential to provide detailed information on the vessel wall, such as plaque composition, elastic wall properties, and even biochemical processes that take place in the plaque. In addition, dynamic and perfusion imaging can provide functional information, e.g. for determining the perfusion or motion of the heart, or to study tumor activity. Owing to the growing complexity and sheer size of cardiovascular data, in combination with the large increase in the number of studies in clinical practice and biomedical research, there is a strong and increasing interest in robust, automated processing tools to aid in the analysis of these data. This research line aims to develop and evaluate novel image processing techniques for visualization, quantification and integrated analysis of multimodal anatomical and functional cardiovascular imaging data. The research in 2005 has focused on: Vessel enhancement filtering (CTA/MRA). Automated central vessel axis extraction in CTA data. Deblurring for improved visualization and calcification quantification in CTA.


78

Quantitative analysis of microvasculature in cancer tissue. Cellular and Molecular Image Analysis Advances in fluorescent probing and microscopic imaging have revolutionized biology in the past decade and have opened up the door to studying the structure and function of even single molecules. In addition, in vivo molecular imaging is expected to have a large impact in clinical practice, as it will be able to study disease processes at the molecular and cellular level, can be used for therapy, and therapy monitoring. Generally, these imaging studies require the processing and analysis of huge amounts of (3D and 4D) image data, which is at present still done mainly by hand. Manual image analysis is very time consuming (thus costly) and also potentially inaccurate and poorly reproducible. As a result, many biologically and clinically interesting questions are either left unaddressed, or answered with great uncertainty. In an attempt to alleviate this problem, this research line aims to develop automated image analysis technology with specific emphasis on accurate and reproducible analysis of cellular and molecular image sequences. The research in 2005 has focused on: Tracking and analysis of biological molecular dynamics in living cells. Pharmacokinetic modeling of contrast uptake in tumors for anticancer treatment. Neuro Image Analysis MR brain imaging is widely used in basic scientific research and in clinical practice, as it is a technique that non-invasively provides both anatomical and functional information. In order to study brain morphometry and function, and its relation to e.g. disease processes or patient characteristics, often large imaging databases are collected. In this research line, advanced techniques for the automated processing and analysis of such databases are developed. The research in 2005 has focused on: Automated MR brain tissue classification Automated quantification of brain structures Intersubject MR brain registration External projects: “3D Multimodal vascular image analysis for improved diagnosis and therapy” NWO STW VICI grant, 1,25 MEuro: 2006-2010 In this project, methods for the integrated analysis of vascular imaging data obtained with various imaging techniques will be developed and evaluated. The focus is on the analysis of the diseased vessel wall, so as to improve diagnosis of cardiovascular diseases, and to monitor and guide their treatment. Collaboration with Applied Physics, Delft “Aneurist: Integrated Biomedical Informatics for the Management of Cerebral Aneurysms” European grant IST call 4, 420 kEuro: 2006-2009 In this multi-center project, the vertical integration of information related to cerebral aneurysms, from the genetic to the epidemiological level is pursued. Our group is involved in the analysis of the morphodynamics of brain aneurysms. “Quantification of tumor vessel morphology: a tool to monitor treatment”. NWO Mozaiek Grant, 180 kEuro: 2005-2009 In this project techniques will be developed and evaluated for the quantitative analysis of multi-sequence MRI data, in order to better monitor the effects of treatment. “ADVAnCE: Automatic Diagnostic Vascular Analysis of CTA Examinations”. Senter grant, 600 kEuro: 2005-2009. In this project, computer aided diagnosis techniques will be developed to support the analysis of CTA data in clinical practice, to improve the workflow of cardiovascular diagnosis. Collaboration with LKEB-Leiden “MOSAIC: Multiscale Object Segmentation by Analyzing Image Content”, NWO EW, 455 kEuro: 2005-2009. In this project, a hierarchical, model-based approach toward multiple-object segmentation is developed, with an application to automated MR brain segmentation. “Model-driven spatiotemporal tracking for quantitative analysis of subcellular dynamics” NWO EW VIDI grant, 600 kEuro: 2005-2009 In this project, automated image analysis techniques for the accurate and reproducible quantification of the motion of subcellular structures from time-lapse fluorescence microscopy image data are developed and evaluated. Collaboration with Applied Physics, Delft Contributions to Books Vos, F.M., R.E. van Gelder, I.W.O. Serlie, J. Florie, Y.C. Nio, F.H. Post, R. Truyen, A.M. Vossepoel, and J. Stoker; Virtual Colonoscopy (Ch.9), in: J. Dankelman, C.A. Grimbergen, H.G. Stassen (eds.), Engineering for Patient Safety: Issues in Minimally Invasive Procedures, Lawrence Erlbaum Associates, Mahwah, NJ, 2005, 206-225.


Papers in international journals De Bruin PW, Dercksen VJ, Post FH, Vossepoel AM, Streekstra, GJ, Vos FM. Interactive 3D segmentation using connected orthogonal contours. Comp in Biol and Med 2005; 34(4):329-346. Letteboer, MM, Willems PW, Viergever MA, Niessen WJ. Brain shift estimation in image-guided neurosurgery using 3-D ultrasound. IEEE Trans Biomed Eng 2005;52(2):268-76. Murphy RF, Meijering E, Danuser G. Special Issue on Molecular and Cellular Bioimaging. IEEE Trans Image Processing 2005;14(9):1233-6. Olabarriaga SD, Rouet JM, Fradkin M, Breeuwer M, Niessen WJ. Segmentation of thrombus in abdominal aortic aneurysms from CTA with nonparametric statistical grey level appearance modeling. IEEE Trans Med Imaging 2005;24(4):477-85. Van de Kraats EB, Carelsen B, Fokkens WJ, Boon SN, Noordhoek N, Niessen WJ, et al. Direct navigation on 3D rotational x-ray data acquired with a mobile propeller C-arm: accuracy and application in functional endoscopic sinus surgery. Phys Med Biol 2005;50(24):5769-81. Van de Kraats EB, Penney GP, Tomazevic D, van Walsum T, Niessen WJ. Standardized evaluation methodology for 2-D-3-D registration. IEEE Trans Med Imaging 2005;24(9):1177-89. Van Dijk EJ, Prins ND, Vermeer SE, Vrooman HA, Hofman A, Koudstaal PJ, et al. C-reactive protein and cerebral small-vessel disease: the Rotterdam Scan Study. Circulation 2005;112(6):900-5. Van Vliet M, van Dijke CF, Wielopolski PA, ten Hagen TL, Veenland JF, Preda A, et al. MR angiography of tumor-related vasculature: from the clinic to the micro-environment. Radiographics 2005;25 Suppl 1:S85-97; discussion S97-8. Ver laan JJ, van de Kraats EB, Oner FC, van Walsum T, Niessen WJ, Dhert WJ. The reduction of endplate fractures during balloon vertebroplasty: a detailed radiological analysis of the treatment of burst fractures using pedicle screws, balloon vertebroplasty, and calcium phosphate cement. Spine 2005;30(16):1840-5. Verlaan JJ, van de Kraats EB, Oner FC, van Walsum T, Niessen WJ, Dhert WJ. Bone displacement and the role of longitudinal ligaments during balloon vertebroplasty in traumatic thoracolumbar fractures. Spine 2005;30(16):1832-9. Verlaan JJ, van de Kraats EB, van Walsum T, Dhert WJ, Oner FC, Niessen WJ. Three-dimensional rotational X-ray imaging for spine surgery: a quantitative validation study comparing reconstructed images with corresponding anatomical sections. Spine 2005;30(5):556-61. Van Walsum Th, Baert, SA, Niessen, WJ. Guide wire reconstruction and visualization in 3DRA using monoplane fluoroscopic imaging. IEEE Transactions on Medical Imaging, 24(5):612-623. Contributions to international conference proceedings Rollano-Hijarrubia, E., F. van der Meer, A. van der Lugt, H. Weinans, H.A. Vrooman, A.M. Vossepoel, and R. Stokking; Improving the imaging of calcifications in CT by histogram-based selective deblurring, in: M.J. Flynn (eds.), Physics of Medical Imaging (Proc. Conf. San Diego, CA, USA, Feb.13-15), Proc. SPIE, vol. 5745-09, 2005, 67-78. Suprijanto, M.W. Vogel, F.M. Vos, H.A. Vrooman, and A.M. Vossepoel; Inter-frame Motion Correction for MR Thermometry, in: J. Duncan, G. Gerig (eds.), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005 (Proc. 8th Int. Conf., Palm Springs, CA, USA, Oct. 26-29), Lecture Notes in Computer Science, vol. 3749, Springer Verlag, Heidelberg, 2005, 580-588. Collaboration Applied Physics, TU Delft. van de Kraats EB, Penney GP, van Walsum Th, Niessen WJ. Multispectral MR to X-Ray Registration of Vertebral Bodies by Generating CT-Like Data. In: Medical Image Computing and Computer-Assisted Intervention - MICCAI 2005. Duncan J, Gerig G (eds.), vol. 3750 of Lecture Notes in Computer Science, Springer-Verlag, Berlin, October 2005, pp. 911-918. Manniesing R, Niessen WJ. Multiscale Vessel Enhancing Diffusion in CT Angiography Noise Filtering. In: Information Processing in Medical Imaging - IPMI 2005. Christensen GE, Sonka M (eds.), vol. 3565 of Lecture Notes in Computer Science, Springer-Verlag, Berlin, July 2005, pp. 138-149.


80

2.3.10 Contribution of UU-WI-ics Geometry, Imaging and Virtual Environments The advances in information and communication technology have led to an enormous amount of spatial and image data. There are many application domains where one needs to analyze, process, or visualize such data. Examples include computer-aided training and simulation, multimedia, medical imaging, molecular modeling, automated cartography, 3D web design, virtual reality, and computer games. The research program is concerned with the design, analysis, and implementation of algorithms and data structures for spatial data, and their application in the above mentioned domains. The type of research ranges from theoretical algorithms research leading to the design and analysis of fundamental geometric algorithms and data structures, to application-oriented research leading to new systems applying algorithmic results in domains like geographic information systems, automated cartography, navigation and manipulation for robots and virtual characters, and content-based image retrieval and multimedia. The program has three main lines of research: Computational geometry The field of computational geometry studies fundamental algorithmic techniques and data structures for geometric problems. The main goal is to design correct and provably efficient algorithms and data structures for generic algorithmic questions dealing with collections of points, lines, polygons, and more complicated geometric primitives in the plane and in higher-dimensional space. The robust implementation of such algorithms has become another key research question in this line of research. We also study geometric algorithmic questions that are derived from particular application domains, in particular from geographic information systems (dealing with terrain data), automated cartography (for example map labeling), robotics, and manufacturing. Imaging and multimedia In this line of research we study algorithmic questions related to the interpretation of images, 3D models, music, and video. The topics we work on include segmentation, feature extraction, matching of objects and organization of media in index structures, with a focus on shape-based multimedia, and music retrieval. The research is concerned with the algorithmic aspects of shape analysis, in particular the representation, decomposition, approximation, and deformation of shapes, the transformation of one shape into another, measures for similarity of shapes, and the organization of shapes in search structures for efficient content-based retrieval. Here shape can be derived from images but also from 3D models and music. Virtual environments Navigation (or motion) and interaction (often in the form of manipulation) are key issues in virtual environments. The research concentrates on algorithms for planning and simulation of motion and manipulation in complex environments, such as models of industrial installations and computer games. Path planning and collision detection are crucial underlying algorithmic techniques for this. Motion is often not a goal in itself but a means of reaching a target area where a manipulation task is to be performed, such as the assembly of a spare part in a virtual factory or the removal of a tissue sample in virtual minimal-invasive surgery. For realistic simulation of these tasks we apply knowledge of techniques from robotics such as kinematics and the mechanics of manipulation, in combination with sophisticated models of physical behavior. In all three lines of research we aim both for new fundamental algorithms and data structures, and for techniques that have been verified on practical problem instances and can be used in the different application domains. External Projects MOCCAM 1999-2005, NWO, 250 KEuro. This project aims at developing new algorithms for motion planning in complex environment. On one hand we concentrate on efficient data structures for storing such environments, and on the other hand, we study probabilistic planners to better understand their behavior, improve them, and extend them to new domains. MOVE-ME 2002-2007, NWO, 150 KEuro. This project studies motion planning for large groups of entities in huge virtual environments. We will try to combine flocking techniques with motion planning techniques to let groups move together in realistic ways. MOVIE 2003-2005, EU Fifth Framework, Utrecht part: 300 Keuro. This project aims at applying motion planning technology to efficiently create natural looking motion for individual entities and groups of entities in complex virtual environment, for example in computer games. The project is a combination of algorithms research to obtain new motion planning techniques and more development work to integrate it in virtual environment applications. Clamp 2003-2007, NWO, 280 KEuro This project studies manipulation problems from an algorithmic perspective. The goal is to devise new algorithms to fixtures 2- and 3-dimensional objects and to reorient those using passive devices.


BRICKS 2004-2008, subproject Modelling, Simulation and Visualization, 320 KEuro Our subproject concentrates on interaction with virtual environments. In particular we work on navigation through virtual environments and manipulation of objects in virtual environments. ASCI groups in Delft and CWI are involved in this project as well. AIM@SHAPE 2004-2007, EU Sixth Framework, Utrecht part: 450 Keuro. This project aims to develop techniques for advanced modeling and analysis of shapes. Within the network we work on merging of scans, hole filling, 3D face recognition, and 3D shape matching and retrieval. SHAME 2000-2005, STW, 450Keuro The goal of this project is to create a SHApe Matchign Environment consisting of a software library of efficient shape matching algorithms based on shape. To achieve this we implement, test, and improve existing shape matching algorithms, and develop new ones. C-Minor 2005-2008, subproject of bsik project MultimediaN, 200 KEuro The objectives of this project is to perform music retrieval on the basis of perceptual and cognitive relevant features, to match these feature efficiently, and to model user feedback. Profi 2005-2007, EU Sixth Framework, 300 KEuro In this collaborative project we aim to invent and develop new techniques for the retrieval of figurative images (such as clip art, logos, signs) from large databases. Our techniques will be based on the extraction and matching of perceptually relevant shape features. GOGO 2005-2009, NWO, 350 KEuro This project is about geometric computation for cartography, realistic terrain modeling, and delineating of imprecise geographic regions, where several geometric criteria must be met simultaneously. Doctoral Degrees Tanase, M.; February 16 2005; Shape decomposition and retrieval, Utrecht University, Utrecht, 169 pp.,Project: MindShade, 2e geldstroom. Giannopoulos, P.; September 21 2005; Geometric matching of weighted point sets, Utrecht University, Utrecht, 129 pp., Project: SHAME, 2e geldstroom. Books Erdman, M., Hsu, D., Overmars, M.H., & Stappen, A.F. van der (Eds.). (2005). Algorithmic Foundations of Robotics VI (Springer Tracts in Advanced Robotics, 17). Berlin-Heidelberg: Springer-Verlag. Merabti, M., Lee, N., Overmars, M.H., & El Rhalibi, A. (Eds.). (2005). Proceedings Third International Game Design and Technology Workshop and Conference. Liverpool: Liverpool John Moores University. Contribution to Books Stappen, A.F. van der (2005). Immobilization: analysis, existence, and output-sensitive synthesis. In R. Janardan, M. Smid, & D. Dutta (Eds.), Computer-Aided Design and Manufacturing (AMS-DIMACS Volumes Series, 67) (pp. 165-187). American Mathematical Society. Papers in international journals Berg, J.P. van den, & Overmars, M.H. (2005). Roadmap-based motion planning in dynamic environments. IEEE transactions on robotics and automation, 21, 885-897. Berg, J.P. van den, & Overmars, M.H. (2005). Using workspace information as a guide to non-uniform sampling in probabilistic roadmap planners. The international journal of robotics research, 24, 1055-1071. Berg, M.T. de, Gudmundsson, J. G., Katz, M.J., Levcopoulos, C., Overmars, M.H., & Stappen, A.F. van der (2005). The TSP with neighborhoods of varying size. Journal of Algorithms, 57, 22-36. Bose, P., & Kreveld, M.J. van (2005). Generalizing monotonicity: on recognizing special classes of polygons and polyhedra. Int. J. Comp. Geom. & Appl., 15, 591-608. Cabello, S., Berg, M.T. de, & Kreveld, M.J. van (2005). Schematization of networks. Comput. Geom. Theory & Appl., 30, 223-238.


82

Demaine, E.D., Erickson, J., Hurtado, F., Iacono, J., Langerman, S., Meijer, H., Overmars, M.H., & Whitesides, S. (2005). Separating point sets in polygonal environments. International journal of computational geometry & applications, 403-419. Gudmundsson, J., Haverkort, H.J., & Kreveld, M.J. van (2005). Constrained higher order Delaunay triangulations. Comput. Geom. Theory & Appl., 30, 271-277. Kreveld, M.J. van, Reinbacher, I., Arampatzis, A., & Zwol, R. van (2005). Multi-dimensional scattered ranking methods for geographic information retrieval. GeoInformatica, 9, 61-84. Kuijper, A., Groep, P. van der, Wall, E. van der, & Diest, P.J. van (2005). Expression of HIF-1alpha and its downstream targets in fibroepithelial tumors of the breast. Breast, 7, 808-818. Veltkamp, R.C. (2005). Multimedia Algorithmics. Multimedia Tools and Applications, 27, 187-193. Contributions to international conference proceedings Berg, J.P. van den, Nieuwenhuisen, D., Jaillet, J., & Overmars, M.H. (2005). Creating robust roadmaps for motion planning in changing environments. In IEEE/RSJ International Conference on Intelligent Robots and Systems (pp. 2415-2421). Berg, J.P. van den, & Overmars, M.H. (2005). Prioritized motion planning for multiple robots. In IEEE/RSJ International Conference on Intelligent Robots and Systems (pp. 2217-2222). Berg, M.T. de, Goaoc, X., & Stappen, A.F. van der (2005). A polynomial-time algorithm to design push plans for sensorless parts sorting. In Proceedings of Robotics: Science and Systems (pp. 89-95). Cheong, J.S., & Stappen, A.F. van der (2005). Output-sensitive computation of all form-closure grasps of a semi-algebraic set. In Proceedings of the IEEE International Conference on Robotics and Automation (pp. 784-790). Demaine, E., Erickson, J., Krizanc, D., Meijer, H., Morin, P., Overmars, M.H., & Whitesides, S. (2005). Realizing partitions respecting full and partial order information. In J. Ryan, P. Manyem, K. Sugeng, & M. Miller (Eds.), Proceedings of the 16th Australasian Workshop on Combinatorial Algorithms (pp. 105-113). Florisson, S., Kreveld, M.J. van, & Speckmann, B. (2005). Rectangular cartograms: construction and animation. In Proc. 21st Annu. ACM Sympos. Comput. Geom. (pp. 372-373). Geraerts, R.J., & Overmars, M.H. (2005). Creating small roadmaps for solving motion planning problems. In Proc. 11th IEEE International Conference on Methods and Models in Automation (pp. 531-536). Geraerts, R.J., & Overmars, M.H. (2005). On improving the clearance for robots in high-dimensional configuration spaces. In IEEE/RSJ International Conference on Intelligent Robots and Systems (pp. 4075-4079). Geraerts, R.J., & Overmars, M.H. (2005). On the Analysis and Success of Sampling Based Motion Planning. In Conference of the Advanced School for Computing and Imaging (pp. 313-319). Geraerts, R.J., & Overmars, M.H. (2005). Reachability analysis of sampling based planners. In Proceedings of the IEEE International Conference on Robotics and Automation (pp. 406-412). Goemans, O.C., Levandowski, A., Goldberg, K., & Stappen, A.F. van der (2005). On the design of guillotine traps for vibratory bowl feeders. In Proceedings of the eleventh annual conference of the Advanced School for Imaging and Computing (pp. 320-327). Goemans, O.C., Levandowski, A., Goldberg, K., & Stappen, A.F. van der (2005). On the design of guillotine traps for vibratory bowl feeders. In Proceedings of the IEEE Conference on Automation Science (pp. 79-86). Gulik, R. van, & Vignoli, F. (2005). Visual playlist generation on the artist map. In Proceedings of the Sixth International Conference on Music Information Retrieval (ISMIR) 2005 . London, UK. Haar, F.B. ter, Cignoni, P., Min, P., & Veltkamp, R.C. (2005). A Comparison of Systems and Tools for 3D Scanning. In 3D Digital Imaging and Modeling: Applications of Heritage, Industry, Medicine and Land . Kamphuis, A., Pettre, J., Overmars, M.H., & Laumond, J.P. (2005). Path finding for the animation of walking characters. In Poster Proceedings Eurographics/ACM Symposium on Computer Animation (pp. 8-9). New York: The Association for Computing Machinery, Inc..


Kamphuis, A., Rook, M., & Overmars, M.H. (2005). Tactical Path Finding in Urban Environments. In Proceedings First International Workshop on Crowd Simulation (pp. 51-60). Lausanne: VRLab, EPFL. Klein, O., & Veltkamp, R.C. (2005). Approximation Algorithms for Computing the Earth Movers Distance Under Transformations. In Proceedings 16th Annual Symposium on Algorithms and Computation (ISAAC) (pp. 1019-1028). Springer LNCS 3827. Klein, O., & Veltkamp, R.C. (2005). Approximation Algorithms for the Earth Movers Distance Under Transformations Using Reference Points. In Proceedings 21th European Workshop on Computational Geometry (EWCG) (pp. 53-56). Kok, T. de, Kreveld, M.J. van, & Löffler, M. (2005). Generating realistic terrains with higher-order Delaunay triangulations. In Proc. 13th Europ. Symp. on Algorithms (pp. 343-354). Berlin: Springer. Kok, T. de, Kreveld, M.J. van, & Löffler, M. (2005). Minimizing local minima in terrains with higher-order Delaunay triangulations. In Abstracts of the 21th Europ. Workshop on Comput. Geom. (pp. 115-118). Kreveld, M.J. van, & Schlechter, T. (2005). Automated label placement for groups of islands. In Proc. 22th Int. Cart. Conf. . Kreveld, M.J. van, & Speckmann, B. (2005). Rectangular cartogram computation with sea regions. In Proc. 22st Int. Cartographic Conference . Niemeijer, M., Ginneken, B. van, Haar, F.B. ter, & Abramoff, M.D. (2005). Automatic detection of the optic disc, fovea and vascular arch in digital color photographs of the retina. In W.F. Clocksin, A.W. Fitzgibbon, & P.H.S. Torr (Eds.), Proceedings of the British Machine Vision Conference (pp. 109-118). Nieuwenhuisen, D., Kamphuis, A., Mooijekind, M., & Overmars, M.H. (2005). Automatic Construction of Roadmaps for Path Planning in Games. In Proceedings Advanced School for Computing and Imaging Conference (pp. 153-163). Delft: ASCI. Nieuwenhuisen, D., Stappen, A.F. van der, & Overmars, M.H. (2005). Path planning for pushing a disk using compliance. In IEEE/RSJ International Conference on Intelligent Robots and Systems (pp. 4061-4067). Overmars, M.H. (2005). Path planning for games. In Proceedings Third International Game Design and Technology Workshop and Conference (pp. 29-33). Liverpool: Liverpool John Moores University. Reinbacher, I., Benkert, M., Kreveld, M.J. van, Mitchell, J.S.B., & Wolff, A. (2005). Delineating boundaries for imprecise regions. In Proc. 13th Europ. Symp. on Algorithms (pp. 143-154). Berlin: Springer. Rook, M., & Kamphuis, A. (2005). Path Finding using Tactical Information. In Poster Proc. Eurographics/ACM SIGGRAPH Symposium on Computer Animation (pp. 18-19). New York: The Association for Computing Machinery, Inc.. Ruifrok, A., Scheenstra, A., & Veltkamp, R.C. (2005). A Survey of 3D Face Recognition Methods. In Audio- and Video-based Biometric Person Authentication (AVBPA) (pp. 891-899). Springer. Ruifrok, A., Scheenstra, A., Bijhold, J., & Veltkamp, R.C. (2005). Facial Image Comparison Using 3D Techniques. In Proceedings 2nd International Conference on Reconstruction of Soft Facial Parts (RSFP) . Luchterhand Publishers. Tanase, M., Veltkamp, R.C., & Haverkort, H.J. (2005). Multiple Polyline to Polygon Matching. In Proceedings 16th Annual Symposium on Algorithms and Computation (ISAAC) (pp. 60-70). Springer LNCS 3827. Tanase, M., & Veltkamp, R.C. (2005). Part-based Shape Retrieval. In Proceedings ACM Multimedia (pp. 543-546). Tanase, M., & Veltkamp, R.C. (2005). Part-based Shape Retrieval with Relevance Feedback. In Proceedings International Conference on Multimedia and Expo (ICME) . Typke, R., Hoed, M. den, Nooijer, J. de, Wiering, F., & Veltkamp, R.C. (2005). A Ground Truth For Half A Million Musical Incipits. In Proceedings of the Fifth Dutch-Belgian Information Retrieval Workshop Proceedings of the Fifth Dutch-Belgian Information Retrieval Workshop (pp. 63-70). Typke, R., Hoed, M. den, Nooijer, J. de, Wiering, F., & Veltkamp, R.C. (2005). A Ground Truth For Half A Million Musical Incipits. In Journal of Digital Information Management (pp. 34-39).


84

Typke, R., Wiering, F., & Veltkamp, R.C. (2005). A Survey of Music Information Retrieval Systems. In Joshua.D. Reiss & Geraint.A. Wiggins (Eds.), Proc. 6th International Conference on Music Information Retrieval (pp. 153-160). London: Queen Mary, University of London. Experimental software Geraerts, R.; SAMPLE, System for Advanced Motion Planning Experiments. 2002-2005, C++, Windows. This is software to experimentally compare various motion planning techniques on different scenes. The software is still developed further. Nieuwenhuisen, D.; CALISTO. 2002-2005, C++, Windows. This is a library for visualization and collision checking in 3-dimensional scenes. The software is still developed further. Grigore, O.; SHAME – Shape Matching Environment 2000-2005, C++, Windows. This is a collection of modules for approximation, optimization, and matching of shapes.

2.3.11 Contribution of TUD-LR-frs Reconstruction of building models from aerial photographs and 2D GIS databases The research on the interpretation of the aerial images focuses on knowledge-based 3D reconstruction of buildings. Three-dimensional building models are increasingly required for urban planning, tourism, telecommunication, etc. Since manual 3D processing of aerial images is very time consuming, speeding up the reconstruction by automatic procedures has become a necessity. Our approach relies on combining pairs of stereo images with 2D GIS (Geographic Information System) data and domain knowledge. The domain knowledge includes a building library describing primitive building models (flat roof, gable roof, and hip roof building) since most buildings can be described as an aggregation of these building primitives. In 2002, the performance of the building reconstruction system was evaluated. About 75% of buildings in a suburban environment were successfully reconstructed. This result implies that a significant cost reduction is possible in the production of 3D city models that are currently based on manual measurements. The results were documented in a PhD thesis that was approved by the supervisor in December 2002. Reconstruction of industrial sites from terrestrial laser scanner data and photographs We present a method for 3D reconstruction of industrial sites using a combination of images and point clouds with a motivation of achieving higher levels of automation, precision, and reliability. Recent advances in 3D scanning technologies have made possible rapid and cost-effective acquisition of dense point clouds for 3D reconstructions. As the point clouds provide explicit 3D information they have a much higher potential for the automation of reconstruction. The modelling pipeline in our algorithm starts from point clouds as the main data source for automation. First of all we segment the point cloud using surface smoothness and detect simple objects like planes and cylinders using Hough Transform. This is followed by fitting of CSG objects to a combination of segments. The fitted CAD models are used as registration targets for adding more scans to the project. Additionally, by fitting the projected edges to image gradients we register images to point clouds. (STAR project) Reconstruction of natural sites from terrestrial laser scanner data and photographs The purpose of the research is 3D modelling and reconstruction of (real world) trees on the basis of terrestrial laser scans. To identify the structure of a tree in terms of stem and branches, an algorithm has been designed in 3D voxel space, based on a selection of basic and advanced 2D raster (image) processing algorithms, transferred into the 3D domain. The selection includes filtering, mathematical morphology, skeletonization, connected component labelling and shortest route computation. (Natscan project) Traffic monitoring and modelling using helicopter video sequences Traffic congestion is an important problem in modern society. Better understanding of its causes is needed to be able to more effectively reduce its effects. Congestion appears under a variety of circumstances. Beside the size and amount of roads and the number of vehicles on those roads, the occurrence of traffic congestion is highly dependent upon the behaviour of the individual vehicle drivers. Important factors are reaction speeds, the gap-acceptance and the lane-changing behaviour of the drivers. Due to the lack of detailed data on driver behaviour, contemporary traffic flow theories and models cannot be rigorously calibrated and validated, although there are very strong indications that these models are not able to correctly capture complex driving dynamics during congested or near-congested traffic flow.


External Projects Tracing Congestion Dynamics STW/CITG, Dec. 2004 – Dec. 2008, k€ 178,4 The goal of the proposed research project is to develop a traffic data collection system based on image sequences taken from a helicopter as well as to use this traffic data to calibrate and improve driver behaviour models. Papers in international journals Abbass Malian, Ali Azizi, Frank A. van den Heuvel, Mahmoud Zolfaghari Development of a robust photogrammetric metrology system for monitoring the healing of bedsores Photogrammetric Record 20 (111) september 2005, pp. 241-273 George Vosselman, Patrice Kessels, Ben Gorte. The utilisation of airborne laser scanning for mapping International Journal of Applied Earth Observation and Geoinformation 6 (2005) pp. 177-186 Contributions to international conference proceedings Marco Baars, Peter van Oosterom, Edward Verbree, Ben Gorte Structuring, indexing, querying and visualizing moving objects in a DBMS 7th AIGLE Conference on Geographic Information Science “29 April – 1 May 2004, Heraklion, Greece Parallel Session 3.3 “Database Technology” pp. 251- 260 T. Rabbani, Frank van den Heuvel Efficient hough transform for automatic detection of cylinders in point clouds, ASCI 2005 conference, Het Heijderbos, Heijen, June 8 -10, 2005 T. Rabbani, Frank van den Heuvel Efficient hough transform for automatic detection of cylinders in point clouds, ISPRS Workshop Laser scanning 2005, Enschede, September 12-14, 2005 pp. 60-65, ISSN 1682-1750 T. Rabbani, Frank van den Heuvel Automatic point cloud registration using constrained search for corresponding objects 7th Conference on Optical 3-D Measurement Techniques, October 3-5, 2005, Vienna, Austria Part 1, pp. 177-186 ISBN 3-9501492-2-8 B.G.H. Gorte, F. Karimi Nejadasl, Serge P. Hoogendoorn Outline extraction of a motorway from helicopter image sequence ISPRS, DAGM Vienna Workshop CMRT 05 August 29-30, 2005 pp. 179-184, ISSN 1682-1750 B.G.H. Gorte, N.Pfeifer, S. Oude Elberink Height texture of low vegetation in airborne laser scanner data and its potential for DTM correction IAPRS Vol XXXVI, 3/W3, Proceedings of Laserscanning 2005, Enschede, The Netherlands, 2005 pp.174-179, ISSN 1682-1750 R. Lindenbergh, N.Pfeifer, T. Rabbani Accuracy analysis of the Leica HDS3000 and feasibility of tunnel deformation monitoring IAPRS Vol XXXVI, 3/W3, Proceedings of Laserscanning 2005, Enschede, The Netherlands, 2005 pp. 24-29, ISSN 1682-1750 G. Sithole, G. Vosselman Filtering of airborne laser scanner data based on segmented point clouds IAPRS Vol XXXVI, 3/W3, Proceedings of Laserscanning 2005, Enschede, The Netherlands, 2005 pp.66-71, ISSN 1682-1750

2.3.12 Contribution of TUD-EWI-mm-cgcc Geometric modeling Research on geometric modeling is directed towards feature modeling, which allows modeling of 3D objects with features, i.e. parametric shapes with a functional meaning. The most important research issues are feature validity maintenance, constraint solving, multiple-view feature modeling and conversion, freeform features, mesh generation, and collaborative design. Scientific visualisation Research on visualisation is focused on the development of algorithms and data structures for extracting information from large scientific and engineering data sets, and visualizing this information with computer graphics and virtual reality techniques. Main application areas are flow visualization and medical visualization. Virtual reality Interaction techniques for exploring and manipulating data in 3D virtual environments, and techniques for collaborative visualization. Important applications are in atmospheric research and molecular biophysics.


86

External projects Direct Numerical Simulation of Oil/Water Mixtures Using Front Capturing Techniques NWO-EW Computationjal Science program, 2001-2005, 1 OIO, Ir. B. Vrolijk Joint project with the Fluid Dynamics group (Mechanical Engineering), and the Numerical Mathematics group (EWI). Goal of the project is the development of visualization techniques for very large 3D time-dependent flow datasets; extension of feature extraction and time tracking techniques to visualize the shape of phase fronts in a multi-phase flow, and its development in time. Semantics of families of objects NWO-EW, 2003-2007, 1 OIO, ir. R. van der Meijden Techniques for defining and classifying objects as member of a family of parametric objects. These methods will be based on semantic feature modelling, in which shapes with well-defined semantics and constraints are used to model objects. Of particular interest are techniques to compute the ranges of parameters that correspond to members of a family. Visualization of Cumulus Clouds in Virtual Reality NWO-EW, 2004-2009, 1 OIO, E.J. Griffith MSc. Joint project with the Physics Department on visualiszation of Large-Eddy Simulations (LES) for atmospheric research to help resolve important research questions on cloud dynamics, cloud mixing phenomena, and precipitation. Our contribution is data handling and interactive visualization for very large time-varying data sets from cloud simulations. Architectures for scientific and medical visualization, Bsik VL-e (Virtual Laboratory for e-science) 2004 - 2007, postdoc, dr. C.P. Botha. The aim of this project is to extend the DeVIDE visualization architecture with more general facilities for grid-based computing (resource and data management) and intelligent techniques for semi-automatic network configuration and prototyping for component development. Cooperation with LUMC (Reiber, Lelieveldt) and UvA (Belleman, Olabarriaga). Techniques for Collaborative Visualization in Virtual Environments, Bsik VL-e 2004-2007, post-doc, dr. M. Koutek. Techniques for distributed and collaborative visualization in virtual environments, and 3D interaction; design of software architectures for multi-platform collaborative VR support.Cooperation with CWI (van Liere), VU (Bal) and TUD-PGS (Epema). Interaction techniques for virtual environments, Bsik BRICKS-MSV2, 2004-2008 1 OIO, ir. G. de Haan. Research on generic Virtual Reality software architecture and interaction techniques for navigation, object manipulation, and exploration, to make 3D interaction more comfortable, fast, and effective. Interaction metaphors such as pseudo-haptic visual force feedback are investigated.Cooperation with CWI (van Liere) and UU (van der Stappen). Surface and volume geometry processing for lesion detection and segmentation in virtual colonoscopy Philips Medical Systems, 2004-2008, 1 AIO, Lingxiao Zhao MSc.Research on the determination and use of geometric properties of the inner surface of the human colon for automatic pre-detection of colonic polyps. Invertigation of higher-order geometric properties (curvature) for claculation of distinctive features for polyp detection. Cooperation with TUD-TNW-QI (van Vliet) Integration of design and analysis NWO-EW, 2005-2009, 1 OIO, drs. M. Sypkens Smit.The goal of this project is to develop efficient meshing methods for finite element analysis. For 3D meshes, a method will be developed to only locally update the mesh where the design model has actually been modified. For cases in which it is unfeasible to analyse a full 3D model, methods will be developed to generate meshes for idealised (simplified) models, and to automatically propagate changes in such a model to the design model. Multifield Medical Visualization NWO-VIEW, 2005-2010, 2 OIO, ir. J. Blaas and S. Busking, MSc. An increasing number of medical acquisition and processing techniques are generating large amounts of multi-field data, such as data from different imaging modalities. Many of these datasets are also time-varying, for example in heart motion studies. In this project, we want to transfer knowledge and techniques from closely related fields such as vector and tensor field visualization in scientific visualization, to medical visualization applications.Joint project with TUE-BME (ter Haar-Romeny, Vilanova). Doctoral Degrees Botha, CP (2005, september 12). Techniques and software architectures for medical visualisation and image processing. TUD Technische Universiteit Delft, 180 pag., ISBN 90-8559-094-9. Internal project.


Contributions to Books Vos, FM, Gelder, RE van, Serlie, IWO, Florie, J, Nio, CY, Post, FH, Truyen, R, Vossepoel, AM, & Stoker, J (2005). Virtual colonoscopy (Ch. 9). In J Dankelman, CA Grimbergen, & HG Stassen (Eds.), Engineering for patient safety: issues in minimally invasive procedures (pp. 206-227). Mahwah (NJ), USA: Lawrence Erlbaum Associates (ISBN 0-8058-4905-X). Papers in international journals Bidarra, AR, Madeira, J, Neels, WJ, & Bronsvoort, WF (2005). Efficiency of boundary evaluation for a cellular model. Computer-aided design, (ISSN 0010-4485), 37(12), 1266-1284. Botha, CP, & Post, FH (2005). Improved perspective visibility ordering for object-order volume rendering. The Visual Computer, (ISSN 0178-2789), 21(11), 887-896. Bruin, PW de, Dercksen, V.J., Post, FH, Vossepoel, AM, Streekstra, G.J., & Vos, FM (2005). Interactive 3D segmentation using connected orthogonal contours. Computers in biology and medicine, (ISSN 0010-4825), 35(4), 329-346. Meiden, HA van der, & Bronsvoort, WF (2005). An efficient method to determine the intended solution for a system of geometric constraints. International journal of computational geometry & applications, (ISSN 0218-1959), 15(3), 279-298. Song, Y, Vergeest, JSM, & Bronsvoort, WF (2005). Fitting and manipulating freeform shapes using templates. Journal of Computing and Information Science in Engineering, (ISSN 1530-9827), 5(2), 86-94. Contributions to international conference proceedings Bidarra, AR, Nyirenda, PJ, & Bronsvoort, WF (2005). A feature-based solution to the persistent naming problem. In LA Piegl (Ed.), Computer-aided design and applications, CAD'05 Vol. 2. Computer-aided design and applications, (ISSN 1686-4360) (pp. 517-526). Bangkok, Thailand: CAD Solutions Company Ltd.. Blaas, J, Botha, CP, Peters, B, Vos, FM, & Post, FH (2005). Fast and reproducible fiber bundle selection in DTI visualization. In CT Silva, E Gröller, & H Rushmeier (Eds.), Proceedings IEEE Visualization 2005 (pp. 59-64). Piscataway (NJ), USA: IEEE (ISBN 0-7803-9462-3). Botha, CP, Graaf, T de, Root, R, Wielopolski, P, Schutte, S, Post, FH, Helm, FCT van der, Simonsz, H, & Simonsz, H (2005). Time-varying three-dimensional vector field visualisation for the analysis of retrobulbar fat mobility during eye motion. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 - Proceedings of the 11th annual conference of the Advanced School for Computing and Imaging (pp. 271-275). Delft, The Netherlands: Advanced School for Computing and Imaging (ISBN 90-803086-0-9). Carpentier, GJP, & Bidarra, AR (2005). Behavioral assumption-based prediction for high-latency hiding in mobile games. In Q Mehdi & N Gough (Eds.), Proceedings of CGAMES'2005 (pp. 83-88). Wolverhampton, UK: University of Wolverhampton (ISBN 0-9549016-2-6). Griffith, EJ, Post, FH, Koutek, M, Heus, T, & Jonker, HJJ (2005). Feature tracking in VR for cumulus cloud life-cycle studies. In R Blach & E Kjems (Eds.), Virtual Environments 2005, Eurographics/ ACM SIGGRAPH Symposium Proceedings (pp. 59-64). Aire-la-Ville, Switzerland: Eurographics Association (ISBN 3-905673-21-5). Haan, G de, Koutek, M, & Post, FH (2005). IntenSelect: using dynamic object rating for assisting 3D object selection. In R Blach & E Kjems (Eds.), Virtual Enviroments 2005, Eurographics/ ACM SIGGRAPH Symposium Proceedings (pp. 201-209). Aire-la-Ville, Switzerland: Eurographics Association (ISBN 3-905673-21-5). Meiden, HA van der, & Bronsvoort, WF (2005). A constructive approach to calculate parameter ranges for systems of geometric constraints. In L Kobbelt & V Shapiro (Eds.), Proceedings of SPM 2005, ACM symposium on Solid and Physical Modeling (pp. 135-142). New York, USA: ACM Press (ISBN 1-59593-015-9). Nyirenda, PJ, Bronsvoort, WF, Langerak, TR, Song, Y, & Vergeest, JSM (2005). A generic taxonomy for defining free-form feature classes. In LA Piegl (Ed.), Computer-aided design and applications, CAD'05 Vol. 2. Computer-aided design and applications, (ISSN 1686-4360) (pp. 497-506). Bangkok, Thailand: CAD Solutions Company Ltd.. Vergeest, JSM, Langerak, TR, Song, Y, Wang, C, Bronsvoort, WF, & Nyirenda, PJ (2005). Towards reverse design of freeform shapes. In V Skala (Ed.), WSCG'2005 Short papers proceedings (pp. 165-168). Plzen, Czech Republic: University


88

of West Bohemia (ISBN 80-903100-9-5). Vrolijk, B, & Post, FH (2005). Fast out-of-core isosurface extraction and rendering of time-varying data sets. In BJA Kröse, HJ Bos, EA Hendriks, & JWJ Heijnsdijk (Eds.), ASCI 2005 - Proceedings of the 11th annual conference of the Advanced School for Computing and Imaging (pp. 120-129). Delft, The Netherlands: Advanced School for Computing and Imaging (ASCI) (ISBN 90-803086-0-9). Experimental software C.P. Botha, DeVIDE = Delft Visualisation and Image Processing Development Environment Developed from 2002-present, language: Python / C++, OS: Linux, Windows DeVIDE is a generic software platform created in order to prototype, test and deploy new visualization and image processing algorithms and ideas, mainly intended for medical applications, but also usable for other applications. It provides a highly interactive and flexible development and test environment, and it features integrated functionality of the VTK and ITK toolkits for visualisation and image processing. Highlight G. de Haan, E.J. Griffith, M Koutek and F.H. Post Supportive Interaction and Hybrid Interfaces in Virtual Environments Hybrid user interfaces integrate well-known 2D user interface elements in a 3D virtual environment, and provide a familiar and portable interface across a variety of Virtual Reality systems. We employ Virtual Reality as a means for enhancing interactive scientific visualization and data exploration. We focus on the visualization and control of simulations of physical processes. The Cloud Explorer, illustrated here, is one of our main applications of interest. The goal of this application is to facilitate cumulus cloud life-cycle studies. Large data sets are generated by atmospheric simulations, from which various information modalities and features are extracted in an offline preprocessing phase. The resulting data can be interactively explored in a virtual environment, such as the Virtual Workbench, equipped with tracked stereo glasses, a stylus and a transparent acrylic hand-held panel called the PlexiPad.


2.3.13 Contribution of UvA- FdNWI-isis The main research themes are In the area of computer vision our efforts concentrate on semantic access to images by providing a description of objects regardless the accidental effects that illumination and viewpoint have on the appearance of the objects. As a next type we do similar research in the description of object classes regardless the accidental recording conditions. Specifically, we employ color as an important source of information, as well as texture and local shape. Content-based access of visual data has been a topic of study for some time now. Based on the invariant colour, texture, and shape features, described above, successful methods have been designed to achieve image description at a semantic level. Recently, we have included the recognition of emotion, motion patterns, scenes, and individuals, in addition to objects and object types as important means to access visual data. Computer vision Colour is an important cue in image analysis. Image processing of color images requires special attention as well as new possibilities to see detail where grey value methods cannot. In previous years, local color, geometric and spatial frequency invariants were developed at ISIS. These local features aim at the robust measurement of the color, shape and texture of an object, under the most common accidental viewing conditions. These accidental conditions can be largely characterized by direction of view, the incident light, the color of the light and the other accidents such as the presence of foreground or background cluttering. An invariant representation of the object implies that the various conditions under which the same view of the object may be perceived do not have to be learned. To test the discriminative power of color, geometric and spatial frequency invariants, the ALOI collection was recorded, containing color images of 1000 real-world objects recorded under various illumination directions, and different illumination colors. Another important topic is feature detection by color saliency boosting. There exists a broad agreement that local features (such as depicted in Figure 2) are an efficient tool for object recognition due to their robustness with respect to occlusion and geometrical transformations. The success of local features is greatly dependent on the information content of the selected features. From information theory it is known that the information content of a feature is related to its frequency of occurrence. Features with high information content, so called salient features, are characterized by a low frequency of occurrence. They represent rare events in the world. Another important research result is on color invariance. We proposed invariants which increase discriminative power for object recognition when compared to the visual measurements of which the invariants were composed. More specifically, the illumination intensity invariant W and the color constant shadow and shading invariant N proof to be discriminative when recognizing objects from a data set containing much photometric variation. However, the invariants are shown to be only marginally color constant. The invariant sets outperform SIFT-features extracted from distinctive key points when the object is recorded under various illumination directions or when the object is rotated in 3 dimensions. Combining local invariant features and the incorporation of color are known to be nontrivial problems in computer vision, for which we have provided solutions. From the general theory, methods are derived for industrial colour vision, analysis of colour in documents, colour in microscopy, general images as appearing on the worldwide web. The topic on colour has led a Ph.D. thesis by J. van de Weijer entitled "Color Features and Local Structure in Images" and many publications in journals (International Journal on Computer Vision and IEEE Trans. on Pattern Analysis and Machine Intelligence). IEEE Trans. on PAMI is the most prominent journal in the area of pattern recognition and is among the top three journals in computer science in the world. We aim at expanding the analysis to take on the same invariant representation of texture next, learning to fuse various color invariant models as well as the semantic meaning of colour. Another important aspect in computer vision is Segmentation and Learning and Tracking. Image segmentation is the task of delineating the image of an object from the real world in the digital data array. It is one of the fundamental difficulties of computer vision, easily surpassed by man's superior capabilities. The difficulty resides in the fact that even man cannot rarely give a formal pictorial description why a boundary is positioned at a certain location and that is what a computer needs to perform the job. We contribute a variety of solutions.

First, we found proper solutions to combine derivatives of the different color channels to compute locale image structures such as edges, corners, circles etc. Further, learning methods have been used to select and weight color (invariant) models for discriminatory and robust image feature detection. Finally, we proposed a classification scheme to classify the physical nature of image structures such as shadow, highlight and material edges/corners. Content-based access of visual data Color plays an indispensable role in the attribution of saliency. Postboxes are painted red to facilitate locating them, flowers take on bright colors to attract birds, and traffic signs use highly saturated colors to attract attention of traffic participants. An


90

evaluation of existing features detectors, including both color and luminance-based methods reveals that these methods fail to detect salient features. We have developed powerful methods to improve the saliency, and thereby the information content, of existing feature detectors. To assess the information content of color features, we investigated their frequency of occurrence. To this end, the probability distribution of color-derivatives, which are the most commonly used building blocks for feature detectors, is computed on a dataset of 40.000 images (the Corel dataset). The distribution is given in Figure 2. The axes are in the opponent color space, where the xy-plane indicates the variation in color values, and the z-axis the variation in intensity. Here an interesting phenomena is revealed, which clarifies the failure of classical feature detectors. Variations in the real world are dominated by luminance changes. Classical feature detectors which equally value changes in the color and in the luminance direction are therefore dominated by luminance variations. To improve the saliency of the detected features, we developed a proper method to correct the inhomogeneous distribution of color derivatives. The function which attains this result, called the color saliency boosting function, regulates the amplification of the color variations. The amount of amplification is derived from information theory applied to the distribution of color derivatives. Results of saliency boosting are depicted in Figure 2. After saliency boosting the detectors focus on the regions with higher information content: the commercial panes in the shopping street, and the yellow-jersey and the American flag in the sport picture.

Figure 2. a) Distribution of color derivatives in opponent color space. The ellipsoid-like surfaces indicate derivatives with equal probability. Variations in the luminance direction (z-axis) occur more frequent than in the variations in the color-plane (xy-plane). b) Kalverstraat-image, with in red a Harris detector and in yellow the Harris detector with color saliency boosting. c) Lance-image, with in yellow the affine invariant features Harris-Laplace detector, and in red the same detector after color boosting. In the context of object recognition, proper methods to recognize object classes have been proposed. The first solution is to formulate the recognition task as a graph problem searching for the characteristic geographical arrangements of (possibly missing) parts. The objective function is Bayesian maximum a posteriori estimation, integrating the image likelihood as a posteriori probability of the part detectors. The variability in the arrangement of object parts is captured by a Gaussian distribution after translation normalization. Further, a sparse representation has been proposed for coarse and fine object recognition The topic on learning for object recognition has led a Ph.D. thesis by Pham Viet Thang entitled "Learning Spatial Relations for Object Recognition" and one publication will appear in IEEE Trans. on PAMIi. Further, the topic on learning and image segmentation has lead to several papers in Conference on Computer Vision and Pattern Recognition, the most prestigious conference on the field of compute vision.


Figure 1. The MediaMill Semantic Video Search Engine, showing various advanced interfaces for effective video retrieval. External Projects D: ImIK 2002-2005, IOP, 400 Keuro, Uva-Science-ISIS and VU. The project considers the interactive exploration of multimedia information and knowledge. A computational theory for visual cognition (NWO Veni) 2003-2007, NWO, Uva-Science-ISIS, 200 KEuro. The project aims at developing a computational theory of vision based on invariant representations. MultimediaN 2004-2008, BSIK, 32 MEuro, UvA-Science-ISIS, UU-CS, UT-CS, TUD-CS, CTIT-CS, VU-CS, TNO, NFI, Telematica Inst., V2, Philips, Ilse, LogicaCMG, Waag Society, DBNL, eMAXX, NOS*NSF, Thales, SPSS, Compano, Paradiso, Verenigd Digitaal Erfgoed. The project aims at various forms of multi-media access including e-documents and video. Image retrieval 2001-2005, IOP, Markgraaf, UvA-Science-ISIS. The project aims at analyzing logos, signs to classify them on perceptual similarity. Tracking Sport Video 2002-2005, IOP, Philips CFT, UvA-Science-ISIS. The project aims at tracking of objects in sport video. AMIS NWO, UvA-Science-ISIS, UU-CS, UT-CS, TUD-CS. The project studies multi-media information analysis. Parallel library tools for image processing NWO, TUD-CS, TUD-PH, UvA-ISIS. The project aims at analyzing parallel processes, processing for image processing. MuNCH a NWO-CATCH project for the semantic access to video archives, with Guus Schreiber, VU and Maarten de Rijke, ISLA-ILPS, Beeld&Geluid.


92

Jade a NWO-Glance project for parallel processing of multimedia data with Henri Bal, VU, and CWI. TODS A project with NFI, the forensic lab and ISIS, for forensic data processing, among others on surveillance and crime scene reconstruction. CIOS, ISIS participates in the TNO-center. The Intelligent Sensory Information Systems group was supported by: bsik, IOP-BV, IOP-MMI, EU, NWO, STW, Netherlands Forensic Institute, TNO, Philips, Beeld en Geluid. Doctoral Degrees T.V. Pham, “Learning Spatial Relations for Object Recognition”, PhD. Thesis, University of Amsterdam, 1 February 2005. C.G.M. Snoek, The Authoring Metaphor to Machine Understanding of Multimedia, Ph.D. Thesis, University of Amsterdam, 26 October 2005. J. van de Weijer, “Color Features and Local Structure in Images”, Ph.D. Thesis, University of Amsterdam, 3 March 2005. Books Ip, H., Smeaton, A., Smith, J., Smeulders A. (eds); Proceedings of the International Conference on Multimedia and Expo 2005. IEEE Press, 2005. Sebe, N., Cohen, I., Garg, A., Huang, T.S.; Machine Learning in Computer Vision, Springer Verlag, ISBN 1-4020-3274-9, 2005. Sebe, N., Lew, M.S., Huang, T.S. (ed); Computer Vision in Human-Computer Interaction, HCI/ICCV 2005, Lecture Notes in Computer Science, Volume 3766, Springer-Verlag, ISBN 3-540-29620-4, October 2005. Contributions to Books Geusebroek, J.M.; Early cognitive computer vision. In: Zelkowitz, M.V (ed), Advances in Computers. Elsevier Science, 2005. Sebe, N., Cohen, I., Huang, T.S.; Multimodal Emotion Recognition. In: Handbook of Pattern Recognition and Computer Vision, World Scientific, ISBN 981-256-105-6, 2005. Sebe, N., Tian, Q.; Ranking Metrics and Evaluation Measures. In: Advances in Imaging and Electron Physics, Elsevier, 2005. Papers in international journals Cohen, I., Sebe, N., Cozman, F.G., Gevers, Th., Huang, T.S.; Learning Probabilistic Classifiers for Human-Computer Interaction Applications. ACM Multimedia Systems Journal, volume 10 (6), 2005, pp 484-498. Hoang, M.A., Geusebroek, J.M., Smeulders, A.W.M.; Color texture measurement and segmentation. Signal Processing, Volume 85(2), 2005, pp 265-275. Pham, T.V., Smeulders, A.W.M.; Object recognition with uncertain geometry and uncertain part detection. Computer Vision Image Understanding, volume 99 (2), 2005, pp 241 – 258. Sebe, N., Cohen, I., Cozman, F.G, Huang, T.S.; Learning Probabilistic Classifiers for Human-Computer Interaction Applications. ACM Multimedia Systems, Journal, Volume 10 (6), 2005, pp 484-498. Weijer, J. van de, Gevers, Th., Geusebroek, J.M.; Color edge and corner detection by photometric quasi-invariants. IEEE Trans. Pattern Anal. Machine Intell. Volume 27 (4), 2005, pp.625-630. Wessels, L.F.A., Reinders, M.J.T., Hart, A.A.M., Veenman, C.J., Dai, H., He, T., Veer, L.J. van ‘t; A protocol for building and evaluating predictors of disease state based on microarray data. Bioinformatics, Volume 21 (19), 2005, pp. 3755-3762. Contributions to international conference proceedings Aldershoff, F., Gevers, TH., Prins, P.; Model-assisted object tracking. SPIE, Internet Imaging, San Jose 2005.


Amores, J., Sebe, N., Radeva, P.; Fast Spatial Pattern Discovery Integrating Boosting with Constellations of Contextual Descriptors. IEEE Conference on Computer Vision and Pattern Recognition (CVPR’05), San Diego, USA, June 2005. Blok, M., Danilouchkine, M.G., Veenman, C.J., Admiraal-Behloul, F., Hendriks, E.A., Reiber, J.H.C., Lelieveldt, B.P.F.; Long-Axis Cardiac MRI Contour Detection with Adaptive Virtual Exploring Robot. Functional Imaging and Modeling of the Heart, Barcelona 2005, Lecture Notes in Computer Science 3504 Springer 2005, ISBN 3-540-26161-3. Damen, R., Putten, M.J.A.M. van, Veenman, C.J.; Feature and classifier selection for computer assisted EEG interpretation in Book of Abstracts of the 12th European Congress of Clinical Neurophysiology (ECCN), Stockholm 2005. Geusebroek, J.M.; The stochastic structure of images. In :R. Kimmel, N. Sochen, and J. Weickert, (ed), Proc. 5th Int. Conf. Scale Space and PDE Methods in Comput. Vision (Scale Space 2005), pages 327-338. Springer Verlag (LNCS), 2005. Gemert, J.C. van, Burghouts, G.J., Seinstra, F.J., Geusebroek, J.M.; Color invariant object recognition using entropic graphs. In: Wijk, J.J. van (ed), ASCI’05, Proceedings of the 11th annual conference of the Advanced School for Computing and Imaging, pp. 192-199. Gevers, Th. and Aldershoff, F.; Color Feature Detection and Classification by Learning. IEEE International Conference on Image Processing (ICIP), Genova 2005. Jaimes, A., Nagamine, T., Liu, J., Omura, K., Sebe, N.; Affective Meeting Video Analysis. IEEE International Conference on Multimedia and Expo (ICME’05), Amsterdam, The Netherlands, 2005. Jaimes, A., Sebe, N.; Multimodal Human Computer Interaction: A Survey. IEEE International Workshop on Human-computer Interaction, Beijing, China, 2005. Lehmann, A.S., Pont, S.C., Geusebroek, J.M.; Tree textures: Modern techniques in art-historical context. In: Chantler, M. (ed) Proc. 4th Int. Workshop on Texture Anal. and Synthesis (Texture 2005), Heriot-Watt University, 2005, pp. 43-48 Nguyen, G.P., Worring, M.; Scenario optimization for interactive category search. 7th ACM SIGMM International Workshop on Multimedia Information Retrieval, Singapore 2005. Viet Pham, T., Smeulders, A.W.M., Ruis, S.; Adaboost learning of shape and colour features for object recognition. ICML 2005 Workshop on Machine Learning Techniques for Processing Multimedia Content, 2005. Sebe, N., Cohen, I., Gevers, Th., Huang, T.S.; Multimodal Approaches for Emotion Recognition: A Survey. Internet Imaging VI, SPIE’05, San Jose 2005. Sebe, N., Cohen, I., Huang, T.S., Gevers, Th.; Semi-supervised Face Detection. IEEE Workshop on Learning in Computer Vision and Pattern Recognition, San Diego, USA, June 2005. Sebe, N., Cohen, I., Huang, T.S., Gevers, Th.; Human-Computer Interaction: A Bayesian Network Approach. IEEE International Symposium on Signals, Circuits and Systems, Iasi, Romania, 2005. Sebe, N., Bakker, E., Cohen, I., Gevers, Th., Huang, T.S.; Bimodal Emotion Recognition. 5th International Conference on Methods and Techniques in Behavioral Research, Wageningen 2005, The Netherlands. Pantic, M., Sebe, N., Cohen, J., Huang, T.S.; Affective Multimodal Human-Computer Interaction. ACM Multimedia, Singapore 2005. Putten, M.J.A.M. van, Vliegen, H.R., Veenman, C.J.; Detecting seizure activity from scalp EEG recordings: a comparison of different features and classifiers. Book of Abstracts of the 12th European Congress of Clinical Neurophysiology (ECCN), Stockholm, Sweden, May, 2005. Putten, M.J.A.M. van, Vliegen, H.R., Veenman, C.J.; Feature and classifier selection for computer assisted EEG interpretation. 3rd European Medical and Biological Engineering Conference; IFMBE European Conference on Biomedical Engineering, Prague, Czech Republic, 2005. Seinstra, F.J., Snoek, C.G.M., Koelma, D.C., Geusebroek, J.M., Worring, M.; User Transparent Parallel Processing of the 2004 NIST TRECVID Data Set. 19th International Parallel and Distributed Processing Symposium, Denver 2005 USA. Snoek, C.G.M, Gemert, J.C. van, Geusebroek, J.M., Huurnink, B., Koelma, D.C., Nguyen, G.P., Rooij, O de,


94

Stokman, H.G.M., Gevers, Th.; Selection and Fusion of Color Models for Feature Detection. International Conference on Computer Vision and Pattern Recognition (CVPR05), San Diego 2005. Stokman, H.M.G., Gevers, Th.; Optimal weighting of color channels for discriminative edge detection. SPIE, Internet Imaging, San Jose 2005. Su, K., Tian, Q., Xue, Q., Sebe, N., Ma, J.; Neighborhood Issue in Single-frame Image Super-resolution. IEEE International Conference on Multimedia and Expo (ICME’05), Amsterdam 2005, The Netherlands. Tax, D.M.J., Veenman, C.J.; Tuning the hyperparameter of an AUCoptimized classifier. 17th Belgium-Netherlands Conference on Artificial Intelligence 2005, pp 224-231. Veenman, C.J., Tax, D.M.J.; A Weighted Nearest Mean Classifier for Sparse Subspaces. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Diego 2005, USA. Weijer, J. van de, Gevers, Th.; Boosting Saliency in Color Image Features. International Conference on Computer Vision and Pattern Recognition (CVPR05), San Diego 2005. Weijer, J. van de, Gevers, Th.; Color Constancy based on the Grey-Edge Hypothesis. IEEE International Conference on Image Processing (ICIP), Genova 2005. Zhang, L., Tian, Q., Sebe, N., Ma, J.; Video Object Boundary Reconstruction by 2-pass voting. IEEE International Conference on Multimedia and Expo (ICME’05), Amsterdam 2005, The Netherlands. Experimental software Geusebroek, J.M., Burghouts, G., Smeulders, A.W.M.; The Amsterdam Library of Object Images. ALOI is a color image collection of one-thousand small objects. In order to capture the sensory variation in object recordings, we systematically varied viewing angle, illumination angle, and illumination color for each object, and additionally captured wide-baseline stereo images. We recorded over a hundred images of each object, yielding a total of 110,250 images for the collection. http://www.science.uva.nl/~aloi/ Geusebroek, J.M.; Fast Anisotropic Gauss Filtering software, performs two-dimensional anisotropic Gaussian smoothing and differentiation by recursive filtering techniques. http://www.science.uva.nl/~mark/downloads/ Koelma, D., Poll, E.; Horus: a vision library based on patterns. http://www.science.uva.nl/~koelma/isis/projects/horus.html van Liempt, M., Smeulders, A.W.M, Nguyen, H.T.; Template Tracking Using Color Invariant Pixel Features. Tracking objects robust to photometric variations, such as shadow, shading, and specularities. Handselected objects are followed real-time from either mpg or webcam video stream. http://staff.science.uva.nl/~mliempt/isis/downloads.html Nguyen, G.P., Worring, M.; A demo on similarity based visualization of large image collections. Seinstra, F.J., Koelma, D.; Parallel Horus - 'Proof-of-concept' implementation (partial version) of user transparent parallel image processing functionality as provided in Horus. http://staff.science.uva.nl/~fjseins/ParHorusCode/ Snoek, C.G.M., Worring, M.; Goalgle: A Soccer Video Search Engine, Goalgle is a prototype search engine for soccer video. A fully automatic soccer video analysis system has been developed which analyses soccer broadcasts and makes them accessible by names of players, teams, and highlight events. http://www.goalgle.com Snoek, C.G.M., Worring, M., Geusebroek, J.M., Koelma, D.C., Seinstra, F.J.; Mediamill demo: Semantic Video Search Engine. This system allows for interactive retrieval based on a lexicon of 32 automatically derived semantic concepts. The demo searches a broadcast news archive of 184 hours. http://staff.science.uva.nl/~cgmsnoek/tv/ Snoek, C.G.M., Worring, M.; VIPER: Video Personalizer. VIPER is a prototype search engine for personalized video retrieval over the Internet. The demo searches a broadcast news archive of 184 hours. http://staff.science.uva.nl/~cgmsnoek/viper/ Todoran, L., Worring, M.; Document Ground Truthing and evaluation tool. A tool has been built for assisting users in ground truthing scanned color documents based on layout and logical information. It has been used to create the UvA color document dataset. Tools for evaluating algorithms with respect to the ground truth have been developed as well. Documents and tools will be made available to the research community.


Highlight We applied our real-time facial expression recognition system on still images, notably on many famous paintings. Our result on Mona Lisa was fist mentioned in one of the December issues of the New Scientist and was cited in the media worldwide, see Figure 3. The innovative part was to adapt the system to accommodate static images. Since the original system uses motion displacements from neutral positions of key facial features to classify the facial expression into one of the emotional classes, in the case of images we needed to create a prototypical neutral face. This was obtained as a statistical average of many Caucasian female faces and was adapted to fit Mona Lisa’s face. The obtained results were surprising in the sense that they were not only credible but also showed a blended facial expression on Mona Lisa’s face, see Figure 3.

Figure 3. Facial expression recognition for Mona Lisa: on the right the wireframe model overlayed on the face; on the left the recognized facial expression.

2.3.14 Contribution of UL-LUMC-lkeb Processing, Interpretation and Visualization of Medical Images Contribution of LUMC-lkeb The main goal of the Division of Image Processing is the research, implementation and validation of image processing approaches, which allow the objective and reproducible assessment of objects in medical images. LKEB activities belong to one of the seven main research fields of the LUMC under the headings “Vascular Medicine” and “Neuro-science”. Part of the research involves computer vision research and algorithm development, whereas clinical applications also play an important role. Applications focus on Neuro-imaging, Pulmonology, Orthopaedics, Cardiology and molecular and cellular imaging. In 2005, important research directions were: Statistical shape modeling Statistical shape models are widely used to integrate a-priori knowledge about shape and image appearance into segmentation algorithms. Research at LKEB is directed towards dimensional extension of statistical shape models. A 3D Active Shape Model has been developed, along with 3D and multi-view Active Appearance Models; these models have been applied to segmentation of cardiac MR, CT, echo and X-angiographic data. Apart from segmentation, we are developing statistical shape models for computer-aided diagnosis to detect cardiac shape- and motion abnormalities in MR images for patients with a cardiac infarction, and for quantifying local shape changes caused by brain diseases. We recently investigated shape differences of the brain ventricles in the aging population and Alzheimer disease. Because of the challenging shape of the ventricles, we developed a new method based on growing neural networks to fully automatically model complexly shaped objects.

Multi-agent image processing The major objective of this research is to develop a general and adaptive learning multi-agent image interpretation system, which automatically learns how to interpret (medical) images from examples and user-interactions. The system should be flexible and easy to adapt to changes in patient context, expert preferences, or imaging devices, by the use of both low-level training / optimization and high-level rules. The system will be applied to very difficult segmentation problems in images that cannot be solved with only traditional and/or model-based segmentation methods alone. To this end we are investigating how probabilistic models, reinforcement learning techniques, evolutionary algorithms, high-level (explicit) knowledge and low-level image processing may be integrated into our current multi-agent image interpretation system. The system will first be used for the interpretation of IntraVascular UltraSound (IVUS) and Computed Tomography Angiography (CTA) images.


96

Development of a virtual mobile exploration robot In this project, we address the problem of virtual central navigation in 3D tubular structures. A virtual mobile robot, equipped with a neuro-fuzzy controller, is trained to navigate inside image datasets of tubular structures, keeping a central position; virtual range sensors are used to sense the surrounding walls and to provide input to the controller. Aim of this research is: first the identification of smooth and continuous central paths which are useful in several medical applications: virtual endoscopy, virtual colonoscopy, virtual angioscopy, virtual bronchoscopy, etc. We fully validated the algorithm on synthetic datasets, and performed successful experiments on a colon dataset. Second to train the virtual robot to detect novelties (pathologies) along the path, just like a security agent would do. It will highlight any significant deviation from the normal environment (training using images acquired from healthy subjects), Clinical Image Analysis Applications Much of the research at LKEB is driven by questions from clinical partners. To this end, we are developing algorithms and software for:

• Detection and quantification of pulmonary emphysema in CT Images • Early detection of micro motion of prosthetic implants in bi-plane X-ray images • Fusion of molecular, functional and structural imaging • Automatic analysis of coronary vessels in CT and intravascular ultrasound images • Automatic analysis of coronary and left-ventricular angiograms • Automatic analysis of cardiac funtion in MR and CT patient studies • Automatic analysis of changes in brain structure with ageing and disease • Automatic analysis of vascular MR data • Automatic white matter lesion detection in MR images of the brain • Automatic analysis of quantitative MR images of the brain

External projects CYTTRON 2004-2008 BSIK, with UL-WI-I and TUD-TNW-tn-qi: Development of a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail. SAVAGE 2004-2008, NWO-E, with UL-WI-I: Development of self-learning image processing agents in a multi-agent framework for medical image interpretation. LAISA 2004-2008, STW: Application of multi-agent image processing platform in IVUS and CTA image data. Automated evaluation of vascular MR image data 2004-2007 STW Autonomous virtual robots for medical image exploration 2003-2007 SNN Automatic analysis of 3D echocardiographic data acquired with a fast rotating ultrasound probe 2003-2007, SENTER IOP BV Software Development for the Detection and Assessment of Small Airways Disease in COPD with Multi Slice Computed Tomography, 2003-2007 Netherlands Asthma Foundation Data fusion of different components of MR patient studies 2002-2006 STW Computer-aided diagnosis for cardiac MR data 2002-2007, NWO VIDI Model-based röntgen-stereofotogrammetric analysis of orthopedic implants 2002-2005 STW


Contributions to books A.F. Frangi, W.J. Niessen, M.A. Viergever, B.P.F. Lelieveldt, “A survey of three-dimensional modelling techniques for quantitative functional analysis of cardiac images”, in “Advanced Image Processing in Magnetic Resonance Imaging” (eds. Landini, Positano, Santarelli), pp 267-341, Taylor & Francis, ISBN 0-8247-2542-5, 2005. B.P.F. Lelieveldt, A.F. Frangi, S.C. Mitchell, H.C. van Assen, S. Ordas, J.H.C. Reiber, M. Sonka, “3D active shape and appearance models in cardiac image analysis”, in “Mathematical Models of Computer Vision: The Handbook” (eds N. Paragios, O. Faugeras), pp 471-484, Springer Verlag, ISBN 0387263713, 2005, M.Sonka, D.R. Thedens, B.P.F. Lelieveldt, S.C. Mitchell, R.J. van der Geest, J.H.C. Reiber, “Cardiovascular MR image analysis”, in “Computer vision beyond the visible spectrum”, (Eds B.Bir, I.Pavlidis), pp 193-240, Springer Verlag, Berlin, ISBN: 1-85233-604-8, 2005. F.Admiraal-Behloul, J.H.C Reiber, “Neural Network System Technology and Fuzzy-Rule Extraction in High Dimensional Data sets” .Invited Book Chapter in Cornelius T. Leondes (ed): Intelligent Knowledge-Based Systems: Business and Technology in the New Millennium , Kluwer Academic Press International, ISBN: 1-4020-7829-3, 2005. Papers in international Journals Admiraal-Behloul, F., Heuvel, D. v. d., Olofsen, H., van Osch, M. J. P., van der Grond, J., and van Buchem, M. A. R. J. H. C., "Fully automatic segmentation of white matter hyperintensities in MR images of the elderly," Neuroimage, vol. 28, no. 3, pp. 607-617, Nov.2005. Bakker, ME, Stolk J, Putter H, Shaker SB, Parr DG, Piitulainen E, Russi EW, Dirksen A, Stockley RA, Reiber JH, and Stoel BC. 2005. Variability in Densitometric Assessment of Pulmonary Emphysema with Computed Tomography. Invest Radiol. 40:777-783. J.G. Bosch, F. Nijland, S.C. Mitchell, B.P.F. Lelieveldt, O.Kamp, M.Sonka, J.H.C. Reiber, Computer-Aided Diagnosis via Model-Based Shape Analysis: Automated Classification of Wall Motion Abnormalities in Echocardiograms, Academic Radiology vol. 12(3):358-67, 2005. Box FMA, Geest RJ van der, Rutten MCM, Reiber JHC. The influence of flow, vessel diameter, and non-Newtonion blood viscosity on the wall shear stress in a carotid bifurcation model for unsteady flow. Invest Radiol 2005; 40: 277-94. Danilouchkine MG, Westenberg JJM, Roos A de, Reiber JHC, Lelieveldt BPF. Operator induced variability in cardiovascular MR: left ventricular measurements and their reproducibility. J Cardiovasc Magnetic Resonance 2005; 7: 447-57. Danilouchkine MG, Westenberg JJM, Reiber JHC, Lelieveldt BPF. Accuracy of short-axis cardiac MRI automatically derived from scout acquisitions in free-breathing and breath-holding modes. MAGMA 2005; 18: 7-18. M.G. Danilouchkine, R.J. van der Geest, J.J.M. Westenberg, B.P.F. Lelieveldt, J.H.C. Reiber, “Impact of Inaccurate Positioning and Tilting of Automatically Planned Short-Axis Stack on Quantification of Left Ventricular Dimensions and Function with Cardiovascular Magnetic Resonance” Journal of Magnetic Resonance Imaging, vol 22(6), pp 754-764, 2005. M.S.Dirksen, J.W. Jukema, J.J. Bax, H.J. Lamb, E. Boersma, J.C. Tuinenburg, J. Geleijns, E.E. van der Wal, A. de Roos. Cardiac Multidetector-Row Computed Tomography in Patients with Unstable Angina. Am. J. Cardiol 2005 Feb 15; 95(4):457-61. van Es, A. C., van der Flier, W. M., Admiraal-Behloul, F., Olofsen, H., Bollen, E. L., Middelkoop, H. A., Weverling-Rijnsburger, A. W., Westendorp, R. G., and van Buchem, M. A., "Magnetization transfer imaging of gray and white matter in mild cognitive impairment and Alzheimer's disease," Neurobiol.Aging, Nov.2005. van der Flier, W. M., Middelkoop, H. A., Weverling-Rijnsburger, A. W., Admiraal-Behloul, F., Bollen, E. L., Westendorp, R. G., and van Buchem, M. A., "Neuropsychological correlates of MRI measures in the continuum of cognitive decline at old age," Dement.Geriatr.Cogn Disord., vol. 20, no. 2-3, pp. 82-88, 2005. van der Flier, W. M., van der Vlies, A. E., Weverling-Rijnsburger, A. W., de Boer, N. L., Admiraal-Behloul, F., Bollen, E. L., Westendorp, R. G., van Buchem, M. A., and Middelkoop, H. A., "MRI measures and progression of cognitive decline in nondemented elderly attending a memory clinic," Int.J.Geriatr.Psychiatry, vol. 20, no. 11, pp. 1060-1066, Nov.2005. Janssen JP, Koning G, Koning PJH de, Bosch JG, Tuinenburg JC, Reiber JHC. A new approach to contour detection in X-ray arteriograms. The Wavecontour. Invest Radiol 2005; 40 (8): 514-20.

Formatted: Bullets and Numbering














98

Kaptein BL, Valstar ER, Stoel BC, Rozing PM, Reiber JHC. A new type of model-based Roentgen stereo-photogrammetic analysis for solving the occluded marker problem. J Biomechanics 2005; 38: 2330-4. Marquering HA, Dijkstra J, Koning PJH de, Stoel BC, Reiber JHC. Towards quantitative analysis of coronary CTA. Int J Cardiovasc Imaging 2005; 21: 73-84. Meijler FL, Billette J, Jalife J, Kik MJL, Reiber JHC, Stokhof AA, Westenberg JJM, Wassenaar C, Strackee J. Atrioventricular conduction in mammalian species: Hemodynamic and electrical scaling. Heart Rhythm 2005; 2: 188-96. S.C. Ordas, H.C. van Assen, J. Puentes, B.P.F. Lelieveldt, A.F. Frangi, “Parametric Optimization of a Model Based Segmentation Algorithm for Cardiac MR Image Analysis: A Grid-Computing Approach”, Studies in Health Technology and Informatics, vol 112 / 2005, pp 146-156, 2005. Paelinck BP, de Roos A, Bax JJ, Bosmans JM, van der Geest RJ, Dhondt D, Parizel PM, Vrints CJ, Lamb HJ. Feasibility of tissue magnetic resonance imaging: a pilot study in comparison with tissue Doppler imaging and invasive measurement. J Am Coll Cardiol. 2005;45 (7):1109-1116 van de Sande, MA, Stoel BC, Obermann WR, Lieng JG, and Rozing PM. Quantitative assessment of fatty degeneration in rotator cuff muscles determined with computed tomography. Invest Radiol 2005; 40:313-319. Steens, S. C., Steup-Beekman, G. M., Bosma, G. P., Admiraal-Behloul, F., Olofsen, H., Doornbos, J., Huizinga, T. W., and van Buchem, M. A., "The effect of corticosteroid medication on quantitative MR parameters of the brain," AJNR Am.J.Neuroradiol., vol. 26, no. 10, pp. 2475-2480, Nov.2005. Stralen M van, Bosch JG, Voormolen MM, Burken G van, Krenning BJ, Geuns R-JM, Lancée CT, Jong N de, Reiber JHC. Left ventricular volume estimation in cardiac three-dimensional ultrasound : a semiautomatic border detection approach. Acad Radiol 2005; 12: 1241-9. Üzümcü M, Geest RJ van der, Sonka M, Lamb HJ, Reiber JHC, Lelieveldt BPF. Multiview Active Appearance Models for simultaneous segmentation of cardiac 2- and 4-chamber long-axis magnetic resonance images. Invest Radiol 2005; 40(4): 195-203. de Vries M, de Koning PJ, de Haan MW, Kessels AG, Nelemans PJ, Nijenhuis RJ, Planken RN, Vasbinder GBC, van Engelshoven JMA, van der Geest RJ, Leiner T. Accuracy of Semiautomated Analysis of 3D Contrast-Enhanced Magnetic Resonance Angiography for detection and quantification of aortoiliac stenoses. Invest Radiol 2005;40:495-503. Wesselius, T., Heersema, D. J., Mostert, J. P., Heerings, M., Admiraal-Behloul, F., Talebian, A., van Buchem, M. A., and De Keyser, J., "A randomized crossover study of bee sting therapy for multiple sclerosis," Neurology, vol. 65, no. 11, pp. 1764-1768, Dec.2005. Westenberg JJM, Doornbos J, Versteegh MIM, Bax JJ, Geest RJ van der, Roos A de, Dion RAE, Reiber JHC. Accurate quantitation of regurgitant volume with MRI in patients selected for mitral valve repair. Eur J Cardio-thoracic Surgery 2005; 27: 462-7. Westenberg JJM, Geest RJ van der, Lamb HJ, Versteegh MIM, Braun J, Doornbos J, Roos A de, Wall EE van der, Dion RAE, Reiber JHC, Bax JJ. MRI to evaluate left atrial and ventricular reverse remodeling after restrictive mitral annuloplasty in dilated cardiomyopathy. Circulation 2005; 112 (Suppl I): I-437-42. van Wijk JPH, Castro Cabezas M, de Koning EJP, Rabelink TJ, van der Geest RJ, Hoepelman IM. In vivo evidence of impaired peripheral fatty acid trapping in patients with human immunodeficiency virus-associated lipodystrophy J of Clin Endocrinology Metab. 2005;90:3575-3582. van den Dool SW, Wasser MN, de Fijter JW, Hoekstra J, van der Geest RJ. Functional renal volume: Quantitative analysis at gadolinium-enhanced MR angiography – Feasibility study in healthy potential kidney donors. Radiology 2005;236:189-195. ten Dam VH , Box FMA, de Craen AJM, van den Heuvel DMJ, Bollen ELEM, Murray HM, van Buchem MA, Westendorp RGJ, Blauw GJ. Lack of Effect of Pravastatin on Cerebral Blood Flow or Parenchymal Volume Loss in Elderly at Risk for Vascular Disease. Stroke. 2005;36:1633.

















Contributions to international conference proceedings Abrahamyan L, Schaap JA, Hoekstra AG, Shamonin D, Box FMA, Geest RJ van der, Reiber JHC, Sloot PMA. A problem solving environment for image-based computational hemodynamics. In: Lecture Notes in Computer Science. VS Sunderam, GD van Albada, PMA Sloot et al. (Eds.). Springer-Verlag Berlin Heidelberg 2005; 287-94. H.C. van Assen, M.G. Danilouchkine, A.F. Frangi, S. Ordas, J.J.M. Westenberg, J.H.C. Reiber, B.P.F. Lelieveldt, “SPASM: Segmentation of Sparse and Arbitrarily Oriented Cardiac MRI data using a 3D ASM”, proc. FIMH 2005, Lecture Notes in Computer Science, vol 3504, pp 33-43, 2005. M. Blok, M.G. Danilouchkine, C.J. Veenman, F. Admiraal-Behloul, E.A. Hendriks, J.H.C. Reiber, B.P.F. Lelieveldt, “Long-Axis Cardiac MRI Contour Detection with Adaptive Virtual Robot”, proc. FIMH 2005, Lecture Notes in Computer Science, vol 3504, pp 54-64, 2005. M.G. Danilouchkine, R.J. van der Geest, F. Admiraal-Behloul, B.P.F. Lelieveldt, J.H.C. Reiber, “Automated Quantification of Cardiac Short-Axis Multi-Slice CT Images for Assessment of Left Ventricular Global Function”, proc SPIE medical imaging 2005, pp 463-474. M.G. Danilouchkine, J.J.M. Westenberg, H.C. van Assen, Johan H.C. Reiber, B.P.F. Lelieveldt, “3D Model-Based Approach to Lung Registration and Prediction of Cardiac Motion due to Respiration”, proc. MICCAI, Lecture Notes in Computer Science, vol 3750, pp 751-759, 2005. Ferrarini L, Olofsen H, Reiber JHC, Admiraal-Behloul F. Exploration of 3D medical images of tubular structures: a neurofuzzy controller for virtual centered navigation. Proceedings 11th Annual Conference ASCI, Heijen: 297-304. Ferrarini L, Olofsen H, Buchem MA van, Reiber JHC, Admiraal-Behloul F. Fully automatic shape modelling using growing cell Neural Networks. In: Medical Image Computing and Computer Assisted Intervention - MICCAI 2005: J. Duncan, G. Gerig (Eds.). Springer-Verlag Berlin, 2005: 451-8. Ferrarini L, Olofsen H, Reiber JHC, Admiraal-Behloul F. A neurofuzzy controller for 3D virtual centered navigation in medical images of tubular structures. In: Artificial Neural Networks: Formal models and their Applications – ICANN 2005. W. Duch, J. Kacprzyk, E. Oja, S. Zadrozny (Eds.). Springer-Verlag Berlin 2005: 371-6 C.R. Oost, G. Koning, M. Sonka, J.H.C. Reiber, B.P.F. Lelieveldt, “Automated Segmentation of X-Ray Left Ventricular Angiograms using Multi-View Active Appearance Models”, proc. FIMH 2005, Lecture Notes in Computer Science, vol 3504, pp 23-32, 2005. S. Ordas, H.C. van Assen, L. Boisrobert, M. Laucelli, J. Puente, B.P.F. Lelieveldt and A.F. Frangi, “Statistical Modeling and Segmentation in Cardiac MRI using a Grid Computing Approach”, Lecture Notes in Computer Science, vol 3470, pp 6-15, 2005 Sanderse M, Marquering HA, Hendriks EA, Lugt A van der, Reiber JHC. Automatic initialization algorithm for carotid artery segmentation in CTA images. In: Medical Image Computing and Computer Assisted Intervention - MICCAI 2005: J. Duncan, G. Gerig (Eds.). Lecture Notes in Computer Science, vol 3750, pp 846-53, 2005 Stralen M van, Bosch JG, Voormolen MM, Burken G van, Krenning BJ, Geuns RJM van, Angelié E, Geest RJ van der, Lancée CT, Jong N de, Reiber JHC. Semi-automatic border detection method for left ventricular volume estimation in 4D ultrasound data. Proceedings 11th Annual Conference ASCI, Heijen: 200-7. Suinesiaputra, A.F. Frangi, H.J. Lamb, J.H.C. Reiber, B.P.F. Lelieveldt, “Automatic Prediction of Myocardial Contractility Improvement in Stress MRI using Shape Morphometrics with Independent Component Analysis”, proc. IPMI 2005, Lecture Notes in Computer Science, vol. 3565, pp 321-332, 2005. Experimental software Vessel wall MASS: segmentation and analysis of vessel wall MR images Segmentation component for QLV-CMS: AAM-based segmentation of left ventricular X-ray angiograms Segmentation component for MRA-CMS: segmentation and analysis of MRA datasets SNIPER: quantification and segmentation software for brain MR studies Cooperations within ASCI Reiber JHC, Koning G, Tuinenburg JC, Lansky A, Goedhart B. Quantitative Coronary Arteriography. In: Coronary Radiology. M. Oudkerk (Ed.) Springer-Verlag Berlin 2004: 41-58.
















100

Adame IM, Geest RJ van der, Wasserman BA, Mohamed MA, Reiber JHC, Lelieveldt BPF. Automatic segmentation and plaque characterization in atherosclerotic carotid artery MR images. MAGMA (Magnetic Resonance Materials in Physics, Biology and Medicine) 2004, 16 (5): 227-34. Adame IM, Geest RJ van der, Wasserman BA, Mohamed M, Reiber JHC, Lelieveldt BPF. Automatic plaque characterization and vessel wall segmentation in magnetic resonance images of atherosclerotic carotic arteries. Medical Imaging 2004: Image Processing. J.M. Fitzpatrick, M. Sonka (Eds.). Proceedings of SPIE: vol 5370: 265-73.

2.3.15 Contribution of UG-CS-IS Biologically motivated Image processing and computer vision Models of the visual cortex are developed and used in computer algorithms. This research is relevant for the areas of image processing, computer vision, pattern recognition, visual perception, and computational neuroscience. Our goal is to understand how humans see and deploy principles of natural vision in artificial vision systems. Using facts from neuroscience and visual perception, we build models of visual neurons and use them in computer simulations to obtain insights and derive practical computer vision algorithms. One example is the model of a grating cell that we developed [Petkov and Kruizinga: 1997 Biological Cybernetics 76 (2) 83-96] and used in a texture operator [Kruizinga and Petkov: 1999 IEEE Transactions on Image Processing 8 (10) 1395-1407], [Grigorescu, Petkov and Kruizinga: 2002 IEEE Trans. on Image Processing 11 (10) 1160-1167]. By means of computer simulations we demonstrated that grating cells may play an important role in the disambiguation of edge information in early vision (texture vs. contours). In contrast to traditional texture operators used in image processing, the texture operator derived from a computational model of a grating cell responds only to texture and does not respond to non-textural features, such as object contours. Another example is our model of non-classical receptive filed inhibition, also called surround suppression, in groups of orientation selective neurons [Petkov and Westenberg: 2003 Biological Cybernetics 88 236-246]. This biological mechanism was simulated and applied to a large number of natural images. We demonstrated that the biological role of this inhibitory mechanism is quick pre-attentive detection of object contours and region boundaries in natural images that are rich in texture. We proposed various contour detection algorithms that deploy this mechanism and showed that they are more effective in detecting object contours and region boundaries than traditional computer vision algorithms for edge detection [Grigorescu, Petkov, Westenberg: 2003 IEEE Trans. on Image Processing 12 (7) 729-739], [Grigorescu, Petkov, Westenberg: 2004 Image and Vision Computing 22 609-622]. A new direction in our research that is inspired by psychological research on the human visual system is the development of a method for the evaluation of the robustness of shape recognition algorithms to incompleteness of contours [Ghosh and Petkov, 2005 IEEE Trans. on Pattern Analysis and Machine Intelligence 27 (11) 1793-1804]. Currently we extend our research on the use of Gabor filters to the area of motion analysis. We also develop algorithms that mimic perceptual grouping by humans. One envisaged application is texture analysis. Shape analysis In shape analysis we study geometrical approaches in which a feature point is characterized by the spatial arrangement of other feature points around it. The collection of local geometrical descriptors for the different feature points of an object is used as a shape characteristics of that object. We study the potential of this approach for object segmentation and classification in natural scenes in practical applications such as traffic sign recognition and image database retrieval [Grigorescu and Petkov: 2003 IEEE Trans. on Image Processing 12 (10) 1274-1286]. We also study the robustness of shape recognition algorithms to incomplete contour representations [Ghosh and Petkov, 2005 IEEE Trans. on Pattern Analysis and Machine Intelligence 27 (11) 1793-1804]. On the applications side, we collaborate with researchers from the University of Leon, Spain, in the area of automatic sperm classification. Connected filters and morphological operators Connected filters are a comparatively new field of research within mathematical morphology. They are edge preserving operators which have found use in noise removal, texture analysis, image compression and description, and feature extraction. Research on connected operators in our group entails algorithm development (including parallelization), development of new classes of filters, applications to 2-D and 3-D medical images, and the development of new connectivity measures for these filters for


increased robustness. One line of this research links to visual cortex modelling: developing morphological analogues of texture operators based on models of certain visual cortical cells. It is hoped these morphological counterparts will be an order of magnitude faster, whilst retaining the useful properties of the cortical cell models. Finally, fast visualization based on connected attribute filters is being explored. Segmentation is a core problem in image analysis, and methods based on both simple thresholding methods and more advanced methods such as watersheds and deformable models are being explored. Application areas are many, but the focus lies on biomedical imaging, both macroscopic (MRI, CT) and microscopic. Machine learning and neural networks Neural networks and other adaptive information processing systems can be applied in classification or regression problems. Such systems can be trained from example data which facilitates the implementation of tasks which cannot be formulated as a simple set of rules in a straightforward way. Currently, prototype based learning schemes for distance based classification are in the center of our interest. We aim at a theoretical understanding of such systems in terms of model situations. The obtained insights allow to understand and predict phenomena which are also relevant in real world learning problems. Furthermore, they enable us to systematically develop and improve practical training algorithms. As an important example we have investigated so-called Learning Vector Quantization (LVQ) systems. This heuristic approach to distance based classification is widely used in practice, but lacks through theoretical understanding. Our framework allows for a systematic comparison of training prescriptions in terms of their typical behavior. Applications of LVQ and similar methods, for instance in the area of medical image analysis, serve as a testing ground for theoretical results and newly developed algorithms. We are furthermore studying networks of so-called Integrate-and-Fire units which model the behavior of biological spiking neurons. We aim at understanding of, for instance, masking effects in the first few layers of the visual system. Systems of adaptive information processing, such as neural networks, can be applied in variety of classification or regression problems. Their ability to learn from example data makes it possible to implement tasks which are not easily formulated as a simple set of rules, but for which example data are available. External projects Dynamical Systems Approach to Texture Analysis 2003-2007, NWO, 110 kEuro Generalized Connected Morphological Operators for Robust Shape Extraction 2002-2006, NWO, 150 kEuro. Papers in international journals A. Ghosh and N. Petkov: Robustness of shape descriptors to incomplete contour representations, IEEE Trans. on Pattern Analysis and Machine Intelligence, 27 (11), 2005, 1793-1804. A. Ghosh and N. Petkov: A cognitive evaluation procedure for contour based shape descriptors, Int. Journal of Hybrid Intelligent Systems, 2 (4), 2005, 237-252. A.C. Jalba, M.H.F. Wilkinson, J.B.T.M. Roerdink, M.M. Bayer and S. Juggins. Automatic diatom identification using contour analysis by morphological scale spaces, Mach. Vis. Applic., 16, 2005, 217-228. C. Bunzmann, M. Biehl, R. Urbanczik, Efficient training of multi-layer perceptrons by Principal Component Analysis, Physical Review E 72, 026117 (2005) T. Volkmann, F. Much, M. Biehl, and M. Kotrla, Interplay of strain relaxation and chemically induced diffusion barriers: nanostructure formation in 2D alloys, Surface Science 586: 157-173 (2005) K. De Raedt, H. De Raedt, and K. Michielsen, Deterministic event-based simulation of quantum interference, Comp. Phys. Comm. 171, 19-39 (2005) K. Michielsen, K. De Raedt, and H. De Raedt, Simulation of Quantum Computation: A deterministic event-based approach, J. Comp. Theor. Nanoscience 2, 227 - 239 (2005) H. De Raedt, K. De Raedt, and K. Michielsen, New method to simulate quantum interference using deterministic processes and application to event-based simulation of quantum computation, J. Phys. Soc. Jpn. Suppl. 76, 16 - 25 (2005)


102

H. De Raedt, K. De Raedt, and K. Michielsen, Event-based simulation of single-photon beam splitters and Mach-Zehnder interferometers, Europhys. Lett. 69, 861 - 867 (2005) Contributions to international conference proceedings A. Ghosh, M. Biehl, and B. Hammer, Dynamical analysis of LVQ-type algorithms, in: M. Cottrell (ed.), Proc. of the 5th Workshop on Self-organizing Maps WSOM 2005, Univ. Paris I, 597-594 (2005) M. Biehl, A. Ghosh, and B. Hammer, The dynamics of Learning Vector Quantization, in: M. Verleysen (ed.), Proc. of the European Symposium on Artificial Neural Networks 2005, d-side publishing, 13-18 (2005) M. Biehl, Lattice gas models and Kinetic Monte Carlo simulations of epitaxial growth, in: A. Voigt (ed.), Multiscale Modeling in Epitaxial Growth, International Series of Numerical Mathematics, Birkhaeuser, 149:1-18 (2005) M. Biehl, F. Much, and C. Vey, Off-lattice Kinetic Monte Carlo simulations of heteroepitaxial growth, in: A. Voigt (ed.), Multiscale Modeling in Epitaxial Growth, International Series of Numerical Mathematics Vol. 149, Birkhaeuser, 41-57 (2005) M. Biehl and F. Much, Off-lattice Kinetic Monte Carlo simulations of Stranski-Krastanov-like growth, in: B. Joyce, P. Kelires, A. Naumovets and D.D. Vvedensky (eds.), Quantum Dots: Fundamentals, Applications, and Frontiers, NATO Science Series II: Mathematics, Physics, and Chemistry Vol. 190, Springer (2005) L. Sanchez, N. Petkov and E. Alegre: Classification of boar spermatozoid head images using a model intracellular density distribution, in M. Lazo and A. Sanfeliu (Eds.), Progress in Pattern Recognition, Image Analysis and Applications: Proc. 10th Iberoamerican Congress on Pattern Recognition, CIARP 2005, Lecture Notes in Computer Science, Vol. 3773, (Springer-Verlag Berlin Heidelberg, 2005) pp.154-160. G. Papari and N. Petkov, Algorithm that mimics human perceptual grouping of dot patterns, in M. De Gregorio et al. (Eds.), Proc. First Int. Symp. on Brain, Vision and Artificial Intelligence BVAI 2005, Naples, October 19-21, 2005, Lecture Notes in Computer Science, Vol. 3704 (Springer-Verlag Berlin Heidelberg, 2005) pp.497-506. A. Ghosh and N. Petkov: Incomplete contour representations and shape descriptors: ICR test studies, in M. De Gregorio et al. (Eds.), Proc. First Int. Symp. on Brain, Vision and Artificial Intelligence BVAI 2005, Naples, October 19-21, 2005, Lecture Notes in Computer Science, Vol. 3704 (Springer-Verlag Berlin Heidelberg, 2005) pp.416-425. L. Sanchez, N. Petkov and E. Alegre: Statistical approach to boar semen head classification based on intracellular intensity distribution, in A. Gagalowicz and W. Philips (Eds.), Proc. Int. Conf. on Computer Analysis of Images and Patterns, CAIP 2005, Lecture Notes in Computer Science, Vol. 3691, (Berlin, Heidelberg: Springer-Verlag, 2005) pp.88-95. G.K. Ouzounis and M.H.F. Wilkinson. Countering Oversegmentation in Partitioning-Based Connectivities. Int. Conf. Image Proc. 2005, September 11-14, Genova, Italy, Vol III, pp. 844-847. G.K. Ouzounis and M.H.F. Wilkinson. Second-Order Connected Attribute Filters Using Max-Trees. In: Mathematical Morphology: 40 years on, Proc. Int. Symp. Math. Morph. (ISMM) 2005, pp. 65-74. E.R. Urbach, N.J. Boersma, and M.H.F. Wilkinson. Vector-Attribute Filters. In: Mathematical Morphology: 40 years on, Proc. Int. Symp. Math. Morph. (ISMM) 2005, pp. 95-104. M.H.F. Wilkinson. Attribute-Space Connected Filters. In: Mathematical Morphology: 40 years on, Proc. Int. Symp. Math. Morph. (ISMM) 2005, pp. 85-94. Web-enabled software N. Petkov and M.B. Wieling. Gabor filter for image processing and computer vision, internet enabled implementation, http://mathlabserver.cs.rug.nl N. Petkov and M.B. Wieling. Canny filter for image processing and computer vision, internet enabled implementation, http://mathlabserver.cs.rug.nl N. Petkov and W. T. Visser. Center-surround and dot-pattern selective operators, internet enabled implementation, http://mathlabserver.cs.rug.nl N. M. Hol, F. Kalsbeek and N. Petkov. Grating cell operator, internet enabled implementation, http://mathlabserver.cs.rug.nl


2.4 D: Multimedia Information Systems

2.4.1 Contribution of TUD-mm-ict Image and Video Processing The research on image and video processing in the ICT Group encompasses four topics: Video compression. Compression techniques are essential in variety of Internet applications, television broadcasting, digital cameras, music distribution and consumer digital video applications, such as DVD's and DV-cam cording. Research concentrates on novel compression techniques for applications where the communication medium is error-prone and cases where the device on which the compression algorithm runs is complexity constrained. Computer vision. Focus is on 3D reconstruction of objects from multiple cameras or moving camera(s) for e.g. face recognition, remote handling, 3D editing for 3D teleconferencing. Recent research also considers the problem of gesture recognition. Cultural heritage informatics. Research concerns the application and development of image processing, pattern recognition and artificial intelligence techniques for the benefit of our cultural heritage. The focus is on virtual restoration of paintings as well as on the dating and authenticity of painting and etchings. Restoration. In the area of post-processing the emphasis is on the quality improvement of degraded digital image sequences. An important application is the restoration of historically important (analogue) film and video recordings. Multimedia Retrieval The research efforts in the ICT Group in the field of multimedia content management address the following challenges: - Automating multimedia content indexing and retrieval processes; - Enabling quick, easy and personalized access to multimedia content. We combine our expertise in multimedia signal processing and machine intelligence with state-of-the-art achievements in the fields of "traditional" information retrieval and human perception and aim at bridging the gap between the measurable properties (features) of one hand, by optimizing the way multimedia content is stored, organized, abstracted and represented and, on the other hand, by developing methods for reliably learning user preferences and for filtering, pruning, adapting and delivering multimedia content accordingly. External Projects Non Obtrusive Gesture Recognition 2003-2007, Delft Research Center ICT, 180 k€. In this gesture recognition project we will focus on problems involved in tracking and interpretation of hand movements (gestures) with cameras without using additional obtrusive sensors or markers and how this can be used optimally in different applications. Gestures that can be indicated as conventionalized symbols are predominantly studied in an application where a sign language recognition system will be developed and integrated in an e-learning system to assist the learning process for deaf and hearing-impaired children of Dutch sign language vocabulary. The focus is on detection and tracking of the hand/arm movements and the extraction of relevant features for recognition. DIWAMETRIC (Digital image Watermarks resistant to geoMETRIC attacks) 2001-2004, STW, 270 k€. This project aims at the development of watermarking techniques that are robust against geometric attacks. The secret information embedded in image or video data should survive most common accidental and malicious processing operations such as filtering, compression, D/A and A/D conversion. Cactus 2002-2005, EZ/Freeband K.I., 1.350 M€, TUD-EWI-mm-cgcc, TUD-EWI-st-pds. In this project we concentrate on the interaction between a user and his mobile device and between the mobile device and the network. Specific research questions concern the discovery of available foreign "resources" and the willingness of users and devices to share their own resources. Authenticiity 2004-2007, NWO/TOKEN2000, 300 k€. Authenticity determination and dating of works of art play an important role in the field of art history. In the AUTHENTIC-project it is investigated how existing and new techniques from digital image processing and knowledge discovery (data mining) can be developed and applied in order to support the process of determination of authenticity, dating and the assessment of other characteristics of both graphic art and paintings. Rembrandt PROJECT 2004-2007, NTWO/TOKEN2000, 300 k€. In this project methods for automatic extraction of watermarks and features concerning the paper structure are developed. On the basis of these features, etchings printed on similar papers can be recognized by means of automatic matching procedures. With this questions concerning authenticity and dating can be answered. Since the aim is to construct a database of all the papers used by Dutch artists, specific attention is paid to how to


104

make this database self-organizing and self-learning. This requires the application of advanced techniques from computational intelligence. I-SHARE 2004-2008, EZ BSIK/Freeband, 8 M€, TUD-EWI-st-pds. Sharing resources in virtual communities for storage, communications and processing of multimedia data. The FREEBAND I-SHARE-project investigates the sharing resources in virtual communities, which are (dynamic) groups of nodes that are willing to collaborate for the better of the whole. In particular, I-SHARE considers the sharing of resources for processing, transmitting and storing multimedia streams: - Funtionality sharing. Specific capabilities of terminals embodied in algorithms or, more abstractly, in services are shared to support applications and to form new ones. - Content sharing. This pertains to particular assets related to local hardware. We think of processing power and storage to start with, but specific processing hardware (co-processors) and special purpose hardware are possible as well. Video-at-your-Fingertips 2004-2008, EZ/OC&W BSIK/Multimedian, 2,5 M€. The goal of this project is video content extraction technology suited for practical use in selectec application contexts. The basis of this technology consists of video content analysis algorithms, which include the modeling of the semantic content in terms of signal processing and analysis routines, and the methods of implementing the content models in software or hardware. In particular, Video-at-your-fingertips considers the following research topics: 1. Robust, unconstrained face detection and recognition with applications in security (smart camera's) and media (consumer home video). 2. Human body motion analysis with applications in surveillance (suspicious behavior and aggression detection) and personal health care (revalidation at home). 3. Video content management: The parsing, pruning, abstracting, summarization and classification of video content with a Video Concert Browser as media application. Contributions to Books Lagendijk, R.L., Biemond, J.; Basis methods for image restoration and identification. Bovik, A. (ed.); Handbook of Image & Video Processing, 2nd ed. US/UK, 2005, 1371 pp., Elsevier Academic Press, ISBN 0-12-119792-1, pp. 167-181 Lagendijk, R.L., Roosmalen, P.M.B., Biemond, J., Rares, A., Reinders, M.J.T.: Video enhancement and restoration. Bovik, A. (ed.); Handbook of Image & Video Processing, 2nd ed. US/UK, 2005, 1371 pp., Elsevier Academic Press, ISBN 0-12-119792-1, pp. 275-295 Rajashekar, U., Bovik, A.C., Sage, D., Unser, M., Karam, L.J., Lagendijk, R.L.; Image Processing Education. Bovik, A. (ed.); Handbook of Image & Video Processing, 2nd ed. US/UK, 2005, 1371 pp., Elsevier Academic Press, ISBN 0-12-119792-1, pp. 73-95. Papers in international journals Chen, M., He, Y., Lagendijk, R.L.; A fragile watermark error detection scheme for wireless video communications. IEEE Transactions on Multimedia, IEEE, ISSN 1520-9210, vol. 7, no. 2, pp. 201-211. Eroglu-Erdem, C., Ernst, F., Redert, A., Hendriks, E.A.; Temporal stabilization of video object segmentation for 3D-TV applications. Signal Processing: Image Communication, Elsevier, ISSN 0923-5965, vol. 20, no. 2, pp. 151-167. Hanjalic, A., Xu, L.Q.; Affective video content representation and modeling. IEEE Transactions on Multimedia, IEEE, ISSN 1520-9210, vol. 7, no. 1, pp. 143-154. Hanjalic, A.; Adaptive Extraction of Highlights from a Sport Video Based on Excitement Modeling. IEEE Transactions on Multimedia, IEEE, ISSN 1520-9210, vol. 7, no. 6, pp. 1114-1122. Haratcherev, I., Taal, J.R., Langendoen, K., Lagendijk, R.L., Sips, H.J.; Automatic IEEE 801.11 rate control for streaming applications. Wireless Communications and Mobile Computing, John Wiley & Sons Ltd., ISSN 1530-8669, vol. 5, no. 4, pp. 421-437. Lin, E.T., Eskicioglu, A.M., Lagendijk, R.L., Delp, E.J.; Advances in digital video content protection. Proceedings of the IEEE (Special issue on: Advances in video coding and delivery), IEEE, ISSN 0018-9219, vol. 93, no. 1, pp. 171-183. Rares, A., Reinders, M.J.T., Biemond, J.; Constrained texture restoration. EURASIP Journal on Applied Signal Processing, Hindawi Publishing Corporation, ISSN 1110-8657. vol. 2005, no. 17, pp. 2758-2771.


Rares, A., Reinders, M.J.T., Biemond, J.; Edge-based image restoration. IEEE Transactions on Image Processing, IEEE, ISSN 1057-7149, vol. 14, no. 10, pp. 1454-1468. Yan, W.-Q., Wang, J., Kankanhalli, S.; Analogies based video editing. Multimedia Systems, Springer-Verlag GmbH, ISSN 0942-4962, vol. 11, no. 1, pp. 3-18. Yan, W.-Q., Wang, J., Kankanhalli, S.: Automatic video logo detection and removal. Multimedia Systems, Springer-Verlag GmbH, ISSN 0942-4962, vol. 10, no. 5, pp. 379-391. Contributions to international conference proceedings Blok, M., Danilouchkine, M.G., Veenman, C.J., Admiraal-Behloul, R., Hendriks, E.A., Reiber, J.H.C., Lelieveldt, B.P.F.; Long axis cardiac MRI contour detection with adaptive virtual exploring robot.The 3rd International Conference on Functional Imaging and Modeling of the Heart, Barcelona, 2-4 June 2005, Proceedings of the 3rd International Conference on Functional Imaging and Modeling of the Heart / Lecture Notes in Computer Science, vol. 3504, Frangi, A., Radeva, P., Santos, P. Hernandez, M. (eds.), Berlin 2005, ISBN 3-540-26161-3 / ISSN 0302-9743, Springer Verlag, pp. 54-64. Cai, R., Lu, L., Hanjalic, A.; Unsupervised content discovery in composite audio. The ACM Multimedia Workshops, Singapore, 6-11 November 2005, Proceedings of the ACM Multimedia Workshops, ACM Press, ISBN 1-59593-273-9, pp. 628-637. Doets, P.J.O., Lagendijk, R.L.; Extracting quality parameters for compressed audio from fingerprints. The Sixth International Conference Music Information Retrieval, London, UK, 11-15 September 2005, Proceedings of the Sixth International Conference Music Information Retrieval, Reiss, J.D., Wiggins, G.A. (eds.), London, Queen Mary University, ISBN 0-9551179-0-9, pp. 6. Hanjalic, A., Nesvadba, J., Benois-Pineau, J.; Moving away from narrow-scope solutions in multimedia content analysis. The 2nd European workshop on the integration of knowledge, semantics and digital media technology, London, UK, 29 November – 1 December 2005, Proceedings of the 2nd European workshop on the integration of knowledge, semantics and digital media technology, IEE, ISBN 0-86341-595-4, ISSN 0537-9989, pp. 6. Haratcherev, I., Langendoen, K.G., Lagendijk, R.L., Sips, H.J.; Link adaptation and cross-layer signaling for wireless video-streaming in a shared medium. The 2005 International Conference on Wireless Networks Communications and Mobile Computing, Maui, Hawaii, USA, 13-16 June 2005, Proceedings of the 2005 International Conference on Wireless Networks Communications and Mobile Computing, Baqai, S.M., Uzmi, Z.A. (eds.), IEEE Communications Society, ISBN 0-7803-9306-6, pp. 5. Haratcherev, I., Taal, J.R., Langendoen, K.G., Lagendijk, R.L., Sips, H.J.; Fast 802.11 link adaptation for real-teim video streaming by cross-layer signaling. The IEEE International Symposium on Circuits and Systems, Kobe, Japan, 23-26 May 2005, Proceedings of the IEEE International Symposium on Circuits and Systems, IEEE, ISBN 0-7803-8835-6, pp. 3523-3526. Hollander, R.J.M. den, Hanjalic, A.; Outlier identification in stereo correspondences using quadrics. The British Machine Vision Conference, Oxford, UK, 5-8 September 2005, Proceedings of the British Machine Vision Conference, Clocksin, W., Fitzgibbon, A., Torr, P. (eds.), ISBN 1-901725-29-4, pp. 1-10. Lichtenauer, J.F., Hendriks, E.A., Reinders, M.J.T.; Isophote properties as features for object detection. Proceedings of the Conference on Computer Vision and Pattern Recognition, San Diego, CA, USA, 20-26 July 2006, Proceedings of he Conference on Computer Vision and Pattern Recognition, Schmid, C., Soatto, S., Tomasi, C. (eds.), IEEE Computer Society, pp. 6. Meyer, P.F.A., Westerlaken, R.P., Klein Gunnewiek, R., Lagendijk, R.L.; Distributed source coding of video with non-stationary side-information. The SPIE-conference Visual Communications and Image Processing, Beijing, China, 12-15 July 2005, Proceedings of the SPIE-conference Visual Communications and Image Processing, Li, S., Pereira, F., Shum, H.-Y., Tscher, A.G. (eds.), vol. 5960, SPIE, pp. 857-866. Naci, U., Hanjalic, A.; A unified framework for fast and effective shot transition detection based on analysis of spatiotemporal video data blocks. The 4th International Workshop on Content-Based Multimedia Indexing, Riga, Latvia, 21-23 June 2005, Proceedings of the 4th International Workshop on Content-Based Multimedia Indexing, ISBN 90-74821-707, pp. 1-8. Nesvadba, J., Fonseca, P., Sinitsyn, A., Lange, F. de, Thijssen, M., Kaam, P. van, Liu, H., Leeuwen, R. van, Lukkien, J., Korostelev, A., Ypma, J., Kroon, B., Celik, H., Hanjalic, A., Naci, U., Benois-Pineau, J., With, P. de, Han, J.; Real-time and distributed AV content analysis system for consumer electronics networks. The IEEE International Conference on Multimedia


106

& Expo, Amsterdam, 6-8 July 2006, Proceedings of the IEEE International Conference on Multimedia & Expo, EEE, ISBN 0-7803-9332-5, pp. 1-4. Nesvadba, J., Hanjalic, A., Fonseca, P.M., Hanjalic, A., Kroon, B., Celik, H., Hendriks, E.A.; Towards a real-time and distributed system for face detection, pose estimation and face-related features. The 5th International Conference on Methods and Techniques in Behavioral Research, Wageningen, The Netherlands, 30 August – 2 September 2005, Proceedings of the 5th International Conference on Methods and Techniques in Behavioral Research, Noldus, L.P.J.J., Grieco, F., Loijens, L.W.S., Zimmerman, P.H. (eds.), Wageningen, Noldus Information Technology B.V., ISBN 90-74821-707, pp. 1-4. Pouwelse, J., Slobbe, M. van, Wang, J., Reinders, M.J.T., Sips, H.J.; P2P-based PVR recommendation using friends, taste buddies and superpeers. The Workshop Beyond Personalization 2005, held in conjunction with the 2005 International Conference on Intelligent User Interfaces, San Diego, USA, 9 January 2005, Proceedings of the Workshop Beyond Personalization, Setten, M. van, McNee, S., Konstan, J. (eds.), pp. 66-71. Sanderse, M., Marquering, H.A., Hendriks, E.A., Lugt, A. van der, Reiber, J.C.; Automatic initialization algorithm for carotid artery segmentation in CTA images. The Medical Image Computing and Computer-assisted Intervention-MICCAI 2005: 8th International Conference, Palm Springs, CA, USA, 26-29 October 2005, Proceedings of the Medical Image Computing and Computer-assisted Intervention-MICCAI 2005: 8th International Conference, Part II / Lecture Notes in Computer Science, vol. 3750, Duncan, J.S., Gerig, G. (eds.), Springer-Verlag, ISBN 3-540-29326-4, ISSN 0302-9743, pp. 846-853. Shterev, I.D., Lagendijk, R.L.; Maximum likelihood amplitude scale estimation for quantization-based watermarking in the presence of dither. SPIE-IS&T Electronic Imaging 2005, San Jose, USA, 17-20 January 2005, Proceedings of SPIE-IS&T Electronic Imaging 2005, vol. 5681, Security, Steganography and Watermarking of Multimedia Contents VIII, Delp III, E.J., Wong, P.W. (eds.), SPIE/IS&R, ISBN 0-8194-5654-3, ISSN 0277-786X, pp. 516-527. Shterev, I.D., Lagendijk, R.L.; Maximum likelihood amplitude scale estimation for quantization-based watermarking. The 26th Symposium on Information Theory in the Benelux, Brussels, 19-20 May 2005, Proceedings of the 26th Symposium on Information Theory in the Benelux, Werkgemeenschap Informatie- en Communicatietheorie (WIC), ISBN 90-71048-21-7, pp. 292-300. Taal, J.R., Lagendijk, R.L.; Fair rate allocation of scalable multiple description video for many clients. The SPIE-conference Visual Communications and Image Processing, Beijing, China, 12-15 July 2005, Proceedings of the SPIE-conference Visual Communications and Image Processing, Li, S., Pereira, F., Shum, H.-Y., Tscher, A.G. (eds.), vol. 5960, SPIE, pp. 2173-2183. Taal, J.R., Lagendijk, R.L.; Scalable multiple description video for many clients. The 26th Symposium on Information Theory in the Benelux, Brussels, 19-20 May 2005, Proceedings of the 26th Symposium on Information Theory in the Benelux, Werkgemeenschap Informatie- en Communicatietheorie (WIC), ISBN 90-71048-21-7, pp. 129-136. Wang, J., Reinders, M.J.T., Lagendijk, R.L., Pouwelse, J.; Self-organizing distributed collaborative filtering. The 28th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, Salvador, Brazil, 15-19 August 2005, Proceedings of the 28th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, Marchionini, G., Moffat, A., Tait, J., Baeza-Yates, R., Ziviani, N. (eds.), The Association for Computing Machinery, SIBN 1-59593-034-5, pp. 659-660. Wang, J., Reinders, M.J.T., Pouwelse, J., Lagendijk, R.L.; Wi-Fi walkman: A wireless handhold that shares and recommends music on peer-to-peer networks. SPIE-IS&T Electronic Imaging 2005, San Jose, USA, Proceedings of the SPIE-IS&T Electronic Imaging, Multimedia and Communications II, Sudharsanan, S., Bove Jr., V.M., Panchanathan, S. (eds.), vol. 5863, SPIE/IS&T, pp. 155-163. Wang, J., Shterev, I.D., Lagendijk, R.L.; Two-band amplitude scale estimation for quantization-based watermarking. The 2005 International Symposium on Intelligent Signal Processing and Communication Systems, Hong Kong, 13-16 December 2005, Proceedings of the 2005 International Symposium on Intelligent Signal Processing and Communication Systems, IEEE, ISBN 0-7803-9266-3, pp. 137-140. Westerlaken, R.P., Klein Gunnewiek, R., Lagendijk, R.L.; The role of the virtual channel in distributed source coding of video. The International Conference on Image Processing, Genova, Italy, 11-14 September 2005, Proceedings of the International Conference on Image Processing , IEEE, ISBN 0-7803-9135-7, pp. I-581-584. Westerlaken, R.P., Klein Gunnewiek, R., Lagendijk, R.L.; Turbo-code based wyner-zic video compression. The 26th Symposium on Information Theory in the Benelux, Brussels, 19-20 May 2005, Proceedings of the 26th Symposium


on Information Theory in the Benelux, Werkgemeenschap Informatie- en Communicatietheorie (WIC), ISBN 90-71048-21-7, pp. 113-120. Cooperations within ASCI Blok, M., Danilouchkine, M.G., Veenman, C.J., Admiraal-Behloul, R., Hendriks, E.A., Reiber, J.H.C., Lelieveldt, B.P.F.; Long axis cardiac MRI contour detection with adaptive virtual exploring robot. The 3rd International Conference on Functional Imaging and Modeling of the Heart, Barcelona, 2-4 June 2005, Proceedings of the 3rd International Conference on Functional Imaging and Modeling of the Heart / Lecture Notes in Computer Science, vol. 3504, Frangi, A., Radeva, P., Santos, P. Hernandez, M. (eds.), Berlin 2005, ISBN 3-540-26161-3 / ISSN 0302-9743, Springer Verlag, pp. 54-64. Haratcherev, I., Taal, J.R., Langendoen, K., Lagendijk, R.L., Sips, H.J.; Automatic IEEE 801.11 rate control for streaming applications.Wireless Communications and Mobile Computing, John Wiley & Sons Ltd., ISSN 1530-8669, vol. 5, no. 4, pp. 421-437. Haratcherev, I., Langendoen, K.G., Lagendijk, R.L., Sips, H.J.; Link adaptation and cross-layer signaling for wireless video-streaming in a shared medium. The 2005 International Conference on Wireless Networks Communications and Mobile Computing, Maui, Hawaii, USA, 13-16 June 2005, Proceedings of the 2005 International Conference on Wireless Networks Communications and Mobile Computing, Baqai, S.M., Uzmi, Z.A. (eds.), IEEE Communications Society, ISBN 0-7803-9306-6, pp. 5. Haratcherev, I., Taal, J.R., Langendoen, K.G., Lagendijk, R.L., Sips, H.J.; Fast 802.11 link adaptation for real-teim video streaming by cross-layer signaling. The IEEE International Symposium on Circuits and Systems, Kobe, Japan, 23-26 May 2005, Proceedings of the IEEE International Symposium on Circuits and Systems, IEEE, ISBN 0-7803-8835-6, pp. 3523-3526. Pouwelse, J., Slobbe, M. van, Wang, J., Reinders, M.J.T., Sips, H.J.; P2P-based PVR recommendation using friends, taste buddies and superpeers. The Workshop Beyond Personalization 2005, held in conjunction with the 2005 International Conference on Intelligent User Interfaces, San Diego, USA, 9 January 2005, Proceedings of the Workshop Beyond Personalization, Setten, M. van, McNee, S., Konstan, J. (eds.), pp. 66-71. Sanderse, M., Marquering, H.A., Hendriks, E.A., Lugt, A. van der, Reiber, J.C.; Automatic initialization algorithm for carotid artery segmentation in CTA images. The Medical Image Computing and Computer-assisted Intervention-MICCAI 2005: 8th International Conference, Palm Springs, CA, USA, 26-29 October 2005, Proceedings of the Medical Image Computing and Computer-assisted Intervention-MICCAI 2005: 8th International Conference, Part II / Lecture Notes in Computer Science, vol. 3750, Duncan, J.S., Gerig, G. (eds.), Springer-Verlag, ISBN 3-540-29326-4, ISSN 0302-9743, pp. 846-853.

2.4.2 Contribution of UVA-FdNWI-isis Systems for retrieval of video In the theme systems for retrieval of video, we aim to make multimedia archives accessible. To that end, our research efforts concentrate on automatic semantic indexing and interactive retrieval of multimedia sources. To value the merit of our efforts on high international standards, all research is and will be evaluated within the TRECVID benchmark for multimedia retrieval. To this end, we have developed the MediaMill semantic video search engine which uses a lexicon of detectable concepts in combination with several advanced user interfaces, see Figure 1. The MediaMill system obtained the highest score for the interactive video retrieval task of TRECVID 2005. The topic on retrieval of video has led a Ph.D. thesis by C. Snoek entitled "The Authoring Metaphor to Machine Understanding of Multimedia" and many publications in journals (e.g. IEEE Trans. on Multimedia). We will continue this line of research with a further integration of multi-media sources, a further understanding of the structure of video documents and interactive access to the semantic content of large collections of multi-media information using advanced interfaces. The year 2005 has been a period of change for software engineering from the large complex and abstract Horus system which is now completed to smaller targets systems solving one computer vision task at the time but completely. The Monet database kernel and its modules for image and geo-spatial reasoning are realized in close co-operation with the CWI-database group. In the area of database kernels, an innovative experimental analysis uncovered the lack of performance improvement in database technology over the last decade. The underlying reason is the relative progress in CPU- and RAM-technology, which shows an increasing performance bottleneck. As a result, traditional database solutions use less then a few percent of the available resources. This observation has led to novel techniques to measure the resource waste and new database algorithms to avoid resource stales. A separate topic is the study of parallelism in multimedia processing tasks. The purpose of the research is to anticipate on future generation computer systems while constructing a parallel processing library compatible with Horus. The developed parallel software has been applied in the 2005 TRECVID competition.


108

2006 and beyond With first priority we will increase the effectiveness of our solutions in image retrieval and image search engines by expanding on our experience with learning from image databases. Computational efficiency of image search engines will be increased by the joint development with spatial and extensible databases. This is important as it will open up domains of hundreds of thousands of images, a significant step towards data-mining the content. At the same time, we aim to expand to create full access to multimedia documents. The integration of information from text and pictures is a very interesting topic both scientifically, as it reveals a lot about the nature of information, as well as practically as multimedia documents will be ubiquitous as is the need for their access. The MultimediaN project provides the opportunity to reach this goal with the intended delivery of a large-scale experimentation platform for multimedia information analysis. Concerning color research, we aim at the extraction of invariants from interesting regions of the image. Regions improve on the robustness when compared to strictly local interest point based object recognition. For instance, from color distributions we may derive appearance properties. Further, we aim at exploiting distributions of color edges to derive texture properties. We will concentrate on regions that are interesting from an information theoretical point of view. To increase the specifity of regions, we will incorporate the statistics of regions throughout the ALOI collection. We consider this collection to be a natural starting point for visual cognition. Over the years we have invested in a new, object-oriented software platform for vision. By the end of the year we hope to deliver a first complete system for internal use with an expected life time of 10 years. Further, the available parallel functionality to heterogeneous wide-area Grid systems is projected to be extended in 2005. The main focus is on the development of an efficient and easy-to-use execution model based on so-called Multimedia Grid Services, i.e. high-performance multimedia functionality that can be invoked from within sequential applications running on a standard desktop machine. A key example is our Aibo robot dog, whose video data is being processed at multiple cluster systems all over the globe. This research direction is prioritized with the arrival of the new Distributed ASCI Supercomputer 3 (DAS-3), which is co-financed by the MultimediaN consortium. Contributions to Books Geusebroek, J.M., Burghouts, G.J., Gemert, J.C. van, Smeulders, A.W.M.; Invariant representations to prepare for content based image retrieval from first principles. In: Veltkamp, R. (ed), Trends and Advances in Content-Based Image and Video Retrieval. Springer Verlag, Berlin, 2005. Huijsman, D.P., Sebe, N.; Performance Evaluation in Probabilistic Information Retrieval with Noisy Indexes: The Role of Generality. In: Trends and Advances in Content-Based Image and Video Retrieval, Springer. Schmidt, A.R., Manegold, S., Kersten, M.L.; Storing XML Documents in Databases. In: Rivero, L.C., Doorn, J.H., Ferraggine, V.E.;Encyclopedia of Database Technologies and Applications, Idea Group Publishing, 2005 Windhouwer, M.A., Kersten, M.L.; Digital Media Warehouses. In: Rivero, L.C., Doorn, J.H., Ferraggine, V.E.;Encyclopedia of Database Technologies and Applications, Idea Group Publishing, 2005. Papers in international journals Abiteboul, S., Kersten, M.L., Zdonik S.B.; The Lowell database research selfassessment. Communications ACM, volume 5, 2005, pp111-118. Djeraba, C., Sebe, N., Lew, M.S.; Systems and Architectures for Multimedia Information Retrieval. ACM Multimedia Systems Journal, Volume 10 (6), 2005, pp 457-463. Geusebroek, J.M., Burghouts, G.J., Smeulders, A.W.M.; The Amsterdam library of object images. International Journal Computer Vision, volume 61(1), 2005, pp 103-112. Hollink, L., Nguyen, G., Koelma, D.C., Schreiber, A., Worring, M.; Assessing user behavior in news video retrieval. IEE proceedings on Vision, Image and Signal Processing, Volume 152 (6) 2005, pp. 911-918. Huijsmans, D.P., Sebe, N.; How to Complete Performance Graphs in Content-Based Image Retrieval: Add Generality and Normalize Scope. IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume 27 (2), 2005, pp. 245-251. Nguyen, G.P., Worring, M.; Relevance feedback based saliency adaptation in CBIR. Multimedia Systems, Volume 10 (6), 2005, pp 499-512. Snoek, C.G.M., Worring, M.; Multimodal Video Indexing: A Review of the State-of-the-art. Multimedia Tools and Applications, Volume 25 (1), 2005, pp. 5-35.


Snoek, C.M.G., Worring, M.; Multimedia Event-Based Video Indexing using Time Intervals. IEEE Transactions on Multimedia, Volume 7 (4), 2005, pp.638-647. Contributions to international conference proceedings Amores, J., Sebe, N., Radeva, P.; Efficient Object-class Recognition by Boosting Contextual Information. Iberian Conference on Pattern Recognition and ImageAnalysis, Estoril, Portugal, June 2005. Ballegooij, R. van, Cornacchia, R., Vries, A.P. de, Kersten, M.L.; Distribution Rules for Array Database Queries. International Workshop on Database and Expert Systems Application, Copenhagen 2005, pp 55-64. Hollink, L., Worring, M., Schreiber, A.; Building a Visual Ontology for Video Retrieval. ACM MultiMedia, Singapore 2005. Kersten, M.L., Manegold, S.; Cracking the Database Store. Biennial Conference on Innovative Data Systems Research (CIDR), Asilomar, CA, USA, January 2005. Nguyen, G.P., Worring, M.; Similarity based visualization of image collections. Seventh International Workshop of the EU Network of Excellence DELOS on Audio-visual Content and Information Visualization in Digital Librarires (AVIVDiLib’05), Cortona, Italy, 2005. Sebe, N., Huijsmans, D.P., Tian, Q., Gevers, Th.; Toward Complete Performance Characterization in Content-based Retrieval. Internet Imaging VI, SPIE’05, San Jose 2005. Snoek, C.G.M., Worring, M., Geusebroek, J.M., Koelma, D.C., Seinstra, F.J. On the Surplus Value of Semantic Video Analysis Beyond the Key Frame. IEEE International Conference on Multimedia & Expo (ICME), Amsterdam, The Netherlands, July 2005. Snoek, C.G.M., Koelma D.C., Rest, J. van, N. Schipper, F. J. Seinstra, A. Thean, and M. Worring; MediaMill: Searching Multimedia Archives based on Learned Semantics. IEEE International Conference on Multimedia & Expo (ICME), Amsterdam, The Netherlands, July 2005. Seinstra, F.J., Smeulders, A.W.M., Veenman, C.J., Worring, M.; The MediaMill TRECVID 2005 Semantic Video Search Engine. 3rd TRECVID Workshop, Gaithersburg, USA, November 2005. Snoek, C.G.M., Worring, M., Smeulders, A.W.M. Early versus Late Fusion in Semantic Video Analysis. ACM Multimedia, Singapore 2005. Snoek, C.G.M., Worring, M., Gemert, J.C. van, Geusebroek, J.M., Koelma, D.C., Nguyen, G.P., Rooij, O. de, Seinstra, F.J.; MediaMill: Exploring News Video Archives based on Learned Semantics. ACM Multimedia, Singapore 2005. * Awarded as best technical demonstration at ACM Multimedia 2005. Snoek, C.M.G., Worring, M.; Multimedia Pattern Recognition in Soccer Video using Time Intervals. In: Classification the Ubiquitous Challenge GfKl2004. Series: Studies in Classification, Data Analysis, and Knowledge Organization, Springer-Verlag, 2005. Experimental software Gevers, Th., Smeulders, A.W.M.; ZOMAX, A General Purpose Image Processing and Retrieval System for the World Wide Web. The system provides an interactive image processing module and an image search engine, called PicToSeek, for searching images on the web. http://www.science.uva.nl/research/isis/zomax.html


110

Chapter 3 ASCI-wide Events

3.1 Cooperation within ASCI Distributed ASCI Supercomputer DAS 2 and DAS 3 The DAS 2 will be succeeded by DAS 3: 2005-2009, N W O-EW, 900 KEuro, cooperation with all partners of the ASCI graduate school and SURFnet. A Computer Science Grid with revolutionairy Optical interconnect DAS-3 (Distributed ASCI Supercomputer) is a wide-area distributed system designed by the Advanced School for Computing and Imaging (ASCI). As one of its distinguishing features, DAS-3 employs a very novel internal wide area interconnect based on light paths. The DAS-3 system is used for research on parallel and distributed computing and imaging. DAS-3 consists of 272 compute nodes, spread out over five clusters, located at the four universities: Vrije Universiteit, Amsterdam (VU) Leiden University (LU) University of Amsterdam (UvA) Delft University of Technology (TUD) The MultimediaN project (UvA-MN). The system is currently being built by ClusterVision. Unlike its predecessor, DAS-2, DAS-3 is rather heterogeneous in design. “ADVAnCE: Automatic Diagnostic Vascular Analysis of CTA Examinations”. Senter grant, 600 kEuro: 2005-2009. In this project, computer aided diagnosis techniques will be developed to support the analysis of CTA data in clinical practice, to improve the workflow of cardiovascular diagnosis. EUR-RMI-bigr collaboration with LKEB-Leiden AMIS NWO, UvA-Science-ISIS, UU-CS, UT-CS, TUD-CS. The project studies multi-media information analysis. ARCHER (Architecture Exploration) Period: 2000-2005, Funding: Philips, 0.5 Mfl, UL-WI-i, UvA-FdNWI-caps. Architecture Exploration. The objective is to explore the design space of (lumped) embedded systems for multimedia applications at a high level of abstraction. Archer supports fast, abstract system-level and platform-based exploration and design strategies for the multimedia application domain. Platforms consist of a number of heterogeneous computing components, and a communication, synchronization, and storage infrastructure. The Archer strategy is implemented in software in three layers: An application layer, an architecture layer, and a mapping layer. Applications and Platform are modeled at a high level of abstraction, and the Mapping consists of transformations that take an application (model) representation to an architecture (model) representation. Compaan and Laura are used to calibrate the abstract computing component models in an Archer platform. Archer then concentrates on performance and cost exploration of the platform at the higher level of abstraction. In 2003, Archer has shown its methodology is capable of predicting performance and cost with a confidence level that is well within the range of requirements. Duration: October 2000 – October 2004 Architectures for scientific and medical visualization, Bsik VL-e (Virtual Laboratory for e-science) 2004 - 2007, postdoc, dr. C.P. Botha. The aim of this project is to extend the DeVIDE visualization architecture with more general facilities for grid-based computing (resource and data management) and intelligent techniques for semi-automatic network configuration and prototyping for component development. TUD-EWI-cgcc cooperation with LUMC (Reiber, Lelieveldt) and UvA (Belleman, Olabarriaga). Artemis STW/Progress,2005-2009, Euro 500K is a cooperation between the University of Amsterdam, Leiden University, Delft University of Technology and Philips Research, studies the design of Network-on-Chip (NoC) based MP-SoCs. Our role in this project constitutes of developing a high-level modeling and simulation framework for system-level design space exploration of these NoC-based MP-SoCs. ARTEMISIA 2004-2009, Progress/STW, 930 K Eur. Leiden University, Delft University of Technology, University of Amsterdam, Philips Research Value: 1 Ph.D. student, 1 Post-Doc Sponsor: PROGRESS, Philips Research The project includes the following: Firstly, translate given applications input-output equivalent process networks (an automated translator does exist for a restricted class of applications and is called Compaan. Secondly, model the platform, the architecture templates derived from it as well as the instantiation of these to architectures. Also on a high level of abstraction.Thirdly, provide low-level performance/cost numbers for the processing units on the architecture so that high-level performance/cost numbers are well calibrated and, hence, sufficiently accurate (a calibration tool/platform does exist for


a restricted class processing units and is called Laura. A more advanced calibration tool/platform is to be developed/designed within Artemisia. A structure for maintaining a shared world model in a dynamic environment between differentiated Embedded Systems and their interaction with human supervisors 2001-2006, 490 kf We study embedded autonomous systems in distributed environments, for applications in public safety; monitoring and control of traffic and environmental conditions; assistance and clean-up work in disaster areas. Collaboration between systems requires a collective world model, and we develop methods for its consistency maintenance, in time-critical situations. We use robot soccer (RoboCup) as a case study. This project is a collaboration between UvA and the VU and is sponsored by Progress. Beyond the Ordinary: Design of Embedded Real-time Control (BODERC) 2003-2007. Senter, 2.5 M€, Océ Technologies, Philips CFT, AAS, Imtech ICT, Chess iT, Katholieke Universiteit Nijmegen, Universiteit Twente, TU/e The Boderc project focuses on distributed embedded real-time controllers of complex systems. An Océ printer is taken as a case-study and acts as a driver for the project. The target is an integral approach for a systematic architectural design, modeling, analysis, and validation methodology for such heterogeneous systems. BRICKS, Bsik 2004-2008, subproject Modelling, Simulation and Visualization, 320 KEuro Our subproject concentrates on interaction with virtual environments. In particular we work on navigation through virtual environments and manipulation of objects in virtual environments. UU and other ASCI groups in Delft and CWI are involved in this project as well. Cactus 2002-2005, EZ/Freeband K.I., 1.350 M€, TUD-EWI-ict, TUD-EWI-mm-cgcc, TUD-EWI-st-pds. In this project we concentrate on the interaction between a user and his mobile device and between the mobile device and the network. Specific research questions concern the discovery of available foreign "resources" and the willingness of users and devices to share their own resources. A computational theory for visual cognition (NWO Veni) 2003-2007, NWO, TUD-EWI- cgcc, Uva-Science-ISIS, 200 KEuro. The project aims at developing a computational theory of vision based on invariant representations. CoreGRID 2004-2008, EU (Network of Excellence), EURO 38.000, ASCI partners TUD, VU. This network integrates the grid research of 42 universities in Europe. CIM 2003-2007, Senter, 300k€. TUD-EWI, VU, CWI, Almende, This project aims at the development and specification of distributed incident management techniques. COMBINED systems 2002-2006, 800kE In this project innovative methods for disaster management are being studied. In particular aspects concerning distributed observation systems are investigated. An important property of distributed observation systems is that they autonomously extract information from the monitored area. This information is shared with other services in the disaster management platform. The project is conducted in cooperation with UvA, TUDelft, Thales en TNO within the DECIS laboratory. CSI The CSI project is a close collaboration between LIACS of Leiden University and the CE Laboratory of TU Delft. CSI Media Architecture. The Complex Streamed Instruction Set Architecture (CSI) is a memory-to-memory vector architecture targeted at multimedia applications. A single CSI instruction can process data streams of arbitrary length and, in addition to traditional arithmetic and logical operations, performs data accesses, conversion between storage and computation formats (packing and unpacking), and complex arithmetic hardwired computation. The main new features of the CSI are elimination of the vector sectioning instructions, elimination of the packing/unpacking instructions, and introduction of new complex media related arithmetic instructions. CYTTRON Bio-Computing Search Project 2004-2008 BSIK, Bsik, 8.8 MEuro with UL-LUMC, UL-WI-I and TUD-TNW-tn-qi: Development of a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail.


112

The Cyttron consortium wants to implement a comprehensive, integrated infrastructure for bio-imaging and modeling cells down to atomic detail. We would like to provide a generic tool for identifying the molecular causes of disease, essential for the prevention of disease and the development of new drug and therapies, and to establish a platform for advanced diagnosis and tuning of individualized therapy, increasing effectiveness in health care. D: ImIK 2002-2005, IOP, 400 Keuro, Uva-Science-ISIS and VU. The project considers the interactive exploration of multimedia information and knowledge. “3D Multimodal vascular image analysis for improved diagnosis and therapy” NWO STW VICI grant, 1,25 MEuro: 2006-2010 In this project, methods for the integrated analysis of vascular imaging data obtained with various imaging techniques will be developed and evaluated. The focus is on the analysis of the diseased vessel wall, so as to improve diagnosis of cardiovascular diseases, and to monitor and guide their treatment. EUR-RMI-bigr collaboration with Applied Physics, Delft Fexible Application Mapping Environment (FAME) 2002-2006, NOW-EW, Leiden University, Delft University of Technology, TUE The FAME project aims at developing a novel approach to program optimization, namely, iterative compilation in which the transformation space is searched and profiling is used to measure the impact of transformations. One of the most important goals of this project is to develop heuristics to control the complexity of the search by using analytical models and domain specific knowledge. Freeband/PNP2008 SIK, 500 kEuro (DACS) . Cooperation project between UT and TUDelft, Philips, KPN, WMC,and TNO, to research and develop a Personal Network, focussing on the prototyping of a Personal Mobile Gateway. Gigaport NG Research on Networks, 2005-2007, BSIK, 150KE, UvA & TUD. Within this project the UT investigates self-management of hybrid IP/ Optical networks, and the measurement of traffic flows within such networks. The project is managed by SURFnet, and supports the introduction of SURFnet6. Globule 2003-2007, KEURO 147,(a.o.Proactive Construction of Semantic Overlay Networks) The work on Globule and other peer-to-peer related research is done a cooperation between the VU-WI-i and TUD-EWI-st-pds, TUD-ti-ki. There has been a joint NWO proposal (granted). I-SHARE (part of Freeband) 2004-2008, EZ BSIK/Freeband, 8 M€, TUD-EWI-st-pds. Sharing resources in virtual communities for storage, communications and processing of multimedia data. The FREEBAND I-SHARE-project investigates the sharing resources in virtual communities, which are (dynamic) groups of nodes that are willing to collaborate for the better of the whole. In particular, I-SHARE considers the sharing of resources for processing, transmitting and storing multimedia streams: - Funtionality sharing. Specific capabilities of terminals embodied in algorithms or, more abstractly, in services are shared to support applications and to form new ones. - Content sharing. This pertains to particular assets related to local hardware. We think of processing power and storage to start with, but specific processing hardware (co-processors) and special purpose hardware are possible as well. “Model-driven spatiotemporal tracking for quantitative analysis of subcellular dynamics” NWO EW VIDI grant, 600 kEuro: 2005-2009 In this project, automated image analysis techniques for the accurate and reproducible quantification of the motion of subcellular structures from time-lapse fluorescence microscopy image data are developed and evaluated. EUR- RMI-bigr Collaboration with Applied Physics, Delft Multifield Medical Visualization NWO-VIEW, 2005-2010, 2 OIO, ir. J. Blaas and S. Busking, MSc. An increasing number of medical acquisition and processing techniques are generating large amounts of multi-field data, such as data from different imaging modalities. Many of these datasets are also time-varying, for example in heart motion studies. In this project, we want to transfer knowledge and techniques from closely related fields such as vector and tensor field visualization in scientific visualization, to medical visualization applications. Joint project TUD-ewi-cgcc with TUE-BME (ter Haar-Romeny, Vilanova). MultimediaN 2004-2008, BSIK, 35 MEuro, UvA-Science-ISIS, CWI, TUD-Mediamatica, UT-CTIT-CS, VU-CS, UU-CS, TNO, Nederlands Forensisch Instituut, Telematica Instituut, IBM, V2_, Philips, Ilse Media, LogicaCMG, Waag Society, DBNL, eMAXX,


NOC*NSF, SPSS, Compano, Beeld en Geluid, FabChannel, ZiuZ, DBNL, de Politie Amstellanden, Vereniging Digitaal Erfgoed, Roessingh. The project aims at various forms of multi-media, multi-modal analysis, interaction, system design and knowledge engineering for several scientific and applied science goals. RoboCup/Progress UvA, TUD -TN en VU, also UU (Siks) and TUD- ET (DISC). SCALPE 2004-2008, STW, EURO 220.000, ASCI partners VU and TUD This project deals with high-productivity methods for programming parallel systems on a chip. SmartCam 2002-2006. STW/PROGRESS, TUE-EE-dmes, Philips Natlab, Philips CFT, TNO-FEL, In3D, HP Bristol labs, 1.4 M€, TUD-TNW-tn-ph The SmartCam project investigates low-cost one-chip Smart Camera solutions, contributing to a quantitatively guided design trajectory. In particular, we investigate the impact of current applications, and we try to define relevant architectural parameters and to develop an architectural template. Other aims are to enhance and integrate existing application mapping environments for SIMD and ILP processors. Surface and volume geometry processing for lesion detection and segmentation in virtual colonoscopy Philips Medical Systems, 2004-2008, 1 AIO, Lingxiao Zhao MSc.Research on the determination and use of geometric properties of the inner surface of the human colon for automatic pre-detection of colonic polyps. Invertigation of higher-order geometric properties (curvature) for claculation of distinctive features for polyp detection. TUD-EWI-cgcc Cooperation with TUD-TNW-QI (van Vliet) Techniques for Collaborative Visualization in Virtual Environments, Bsik VL-e 2004-2007, post-doc, dr. M. Koutek. Techniques for distributed and collaborative visualization in virtual environments, and 3D interaction; design of software architectures for multi-platform collaborative VR support. TUD-EWI-cgcc cooperation with CWI (van Liere), VU (Bal) and TUD-PGS (Epema). Token2000 People: A.G. Hoekstra, P.M.A. Sloot, E.V. Zudilova, M. Scarpa, L. Abrahamyan ASCI partner: LKEB/LUMC (P.I.: Reiber) Token2000 is a nationally funded project (NWO), where we collaborate closely with the Universities of Leiden and Twente on the development of an interactive medical application, somewhat similar to the work in CrossGrid. This application is intended for training of surgeons. In collaboration with Leiden University Medical Centre we have created Hemosolve, a problem-solving environment for image based computational Haemodynamics. Hemosolve includes our L-BGK solver, but also a FEM Navier-Stokes solver. Moreover, it contains a 3D editing tool and powerful visualization modules. Project Trader 2004-2008, ESI, 1 AIO, TUD EWI/ UL-LERC/LIACS Modern systems such as household appliances, DVD players, PCs, medical X-ray imaging systems, printers, advanced vehicles, and airplanes rely increasingly on software, in particular for system integration. Embedded software monitors the whole system taking care that the system accomplishes more than its parts would. In such software intensive systems, reliability is of prime importance. Two-level Peer-to-Peer Systems 2003-2007, NWO, KEURO 147, ASCI partners TUD-EWI-pds, VU-group Large-Scale Distributed Systems. This project aims at exploring the notion of superpeers in p2p systems in order to improve the performance of such systems. The work on Globule and other peer-to-peer related research is done a cooperation between the VU-WI-I and TUD-ti-ki. Efforts continue in the Bsik IShare project, in which the TUD and VU jointly work on the Tribler Virtual Laboratory for e-Science (VL-e) 2004-2009, EZ (BSIK program), EURO 1.000.000 (TUD-EWI-st-pds part), ASCI partners TUD-EWI-mm-cgcc, VU-EW-cs-I, UvA-FdNWI-caps, UvA-FdNWI-scs. Amolf, Nikhef, CWI, SARA, IBM, LigicaCMG, Philips, FEI This projects aims at designing and implementing grid technology (schedulers, communication libraries, problem-solving and visualization environments, etc) for virtual laboratories (e.g., for simulations in the sciences) on top of the basic grid fabric. UvA FdNWI-scs In the BSIK funded VL-e project we collaborate with a number of ASCI groups affiliated with VU and UvA VL-e consortium include a number of ASCI members UvA, VU, TUD. few can consider the work developing the generic VL-e middleware as a collaboration of ASCI members


114

In the N W O funded project DIME we collaborate with UL-LUMC-lkeb on a Problem Solving Environment for Computational Hemodynamics, and its potential use in pre-operative planning and training.

3.2 ASCI 2005 Conference ASCI 2005 was the eleventh annual conference organized by ASCI. Apart from keynotes, and paper and poster presentations, the program consisted of theme presentations by senior ASCI researchers. Keynotes

James Crowley from INRIA Grenoble Richard Kleihorst from Philips Research Eindhoven Andrew Herbert from Microsoft Research Cambridge.

About the conference Number of theme presentations 3 Number of poster presentations 23 Numer of paper presentations 32 Date: June 8 - 10, 2005 Location: Conference Het Heijderbos, Heijen Number of participants 130 Proceedings B.J.A. Kröse, H.J. Bos, E.A. Hendriks, Heijnsdijk J.W.J., Proceedings of the eleventh annual conference of the Advanced School for Computing and Imaging, Het Heijderbos, Heijen, June 8-10, 2005, ISBN 90-803086-0-9, Delft, 2005. jms/30-01-07