image understanding & web security henry baird joint work with: richard fateman, allison coates,...
TRANSCRIPT
Image UnderstandingImage Understanding
& Web Security & Web Security
Henry Baird
Joint work with:
Richard Fateman, Allison Coates, Kris Popat,
Monica Chew, Tom Breuel, & Mark Luk
DIAR, Madison, WI – June 21, 2003 (HSB) 2
A fast-emerging research topicA fast-emerging research topic
Human Interactive Proofs (HIPs; definition later):– first instance in 1999– research took hold in CS security theory field first– intersects image understanding, cog sci, etc etc– fast attracting researchers, engineers, & users
This talk:– A brief history of HIPs – Existing systems -- w/ my critiques– Professional activities, so far -- incl. the 1st Int’l Workshop– In detail: PARC’s PessimalPrint & BaffleText
H. Baird & K. Popat, “Web Security & Document Image Analysis,” in J. Hu & A. Antonacopoulos (Eds.), Web Document Analysis, World Scientific, 2003 (in press).
DIAR, Madison, WI – June 21, 2003 (HSB) 3
Straws in the wind…Straws in the wind…
90’s: spammers trolling for email addresses
– in defense, people disguise them, e.g.
“baird AT parc DOT com”
1997: abuse of ‘Add-URL’ feature at AltaVista
– some write programs to add their URL many times
– skewed the search rankings
Andrei Broder et al (then at DEC SRC)
– a user action which is legitimate when performed once becomes abusive when repeated many times
– no effective legal recourse
– how to block or slow down these programs …
DIAR, Madison, WI – June 21, 2003 (HSB) 4
The first known instance…The first known instance…
Altavista’s AddURL filter Altavista’s AddURL filter
1999: “ransom note filter”– randomly pick letters, fonts, rotations – render as an image
– every user is required to read and type it in correctly
– reduced “spam add_URL” by “over 95%”
Weaknesses: isolated chars, filterable noise, affine deformations
M. D. Lillibridge, M. Abadi, K. Bharat, & A. Z. Broder, “Method for Selectively Restricting Access to Computer Systems,” U.S. Patent No. 6,195,698, Filed April 13, 1998, Issued February 27, 2001.
An image of text, not ASCII
DIAR, Madison, WI – June 21, 2003 (HSB) 5
Yahoo!’s “Chat Room Problem”Yahoo!’s “Chat Room Problem”
September 2000
Udi Manber asked Prof. Manuel Blum’s group at CMU:
– programs impersonate people in chat rooms, then hand out ads – ugh!
– how can all machines be denied access to a Web site
without inconveniencing any human users?
I.e., how to distinguish between machines and people on-line
… a kind of ‘Turing test’ !
DIAR, Madison, WI – June 21, 2003 (HSB) 6
Alan Turing (1912-1954)Alan Turing (1912-1954)
1936 a universal model of computation
1940s helped break Enigma (U-boat) cipher
1949 first serious uses of a working computer
including plans to read printed text
(he expected it would be easy)
1950 proposed a test for machine intelligence
DIAR, Madison, WI – June 21, 2003 (HSB) 7
Turing’s Test for AITuring’s Test for AI
How to judge that a machine can ‘think’:
– play an ‘imitation game’ conducted via teletypes
– a human judge & two invisible interlocutors: a human
a machine `pretending’ to be human
– after asking any questions (challenges) he/she
wishes, the judge decides which is human
– failure to decide correctly would be convincing
evidence of machine intelligence (Turing asserted)
Modern GUIs invite richer challenges than teletypes….
A. Turing, “Computing Machinery & Intelligence,” Mind, Vol. 59(236), 1950.
DIAR, Madison, WI – June 21, 2003 (HSB) 8
““CAPTCHAs”:CAPTCHAs”: Completely Automated Public Turing TestsCompletely Automated Public Turing Tests to Tell Computers & Humans Apart to Tell Computers & Humans Apart
challenges can be generated & graded automatically
(i.e. the judge is a machine) accepts virtually all humans, quickly & easily rejects virtually all machines resists automatic attack for many years
(even assuming that its algorithms are known?)
NOTE: the machine administers, but cannot pass the test!
(M. Blum, L. A. von Ahn, J. Langford, et al, CMU-SCS)
L. von Ahn, M. Blum, N.J. Hopper, J. Langford, “CAPTCHA: Using Hard AI Problems For Security,” Proc., EuroCrypt 2003, Warsaw, Poland, May 4-8, 2003 [to appear].
DIAR, Madison, WI – June 21, 2003 (HSB) 9
CMU’s ‘Gimpy’ CAPTCHACMU’s ‘Gimpy’ CAPTCHA
Randomly pick: English words, deformations, occlusions, backgrounds, etc
Challenge user to type in any three of the words Designed by CMU team: tried out by Yahoo! Problem: users hated it --- Yahoo! withdrew it
L. Von Ahn, M. Blum, N. J. Hopper, J. Langford, The CAPTCHA Web Page, http://www.captcha.net.
DIAR, Madison, WI – June 21, 2003 (HSB) 10
Yahoo!’s present CAPTCHA:Yahoo!’s present CAPTCHA: “EZ-Gimpy” “EZ-Gimpy”
Randomly pick: one English word, deformations, degradations, occlusions,
colored backgrounds, etc Better tolerated by users Now used on a large scale to protect various services Weaknesses: a single typeface, English lexicon
DIAR, Madison, WI – June 21, 2003 (HSB) 11
PayPal’s CAPTCHAPayPal’s CAPTCHA
Nothing published Seems to use a single typeface Picks, at random:
letters, overlain pattern Weaknesses: single typeface, simple grid, no image degradations, spaced apart
DIAR, Madison, WI – June 21, 2003 (HSB) 12
Cropping up everywhere… Cropping up everywhere…
In use today, to defend against:– skewing search-engine rankings (Altavista, 1999)– infesting chat rooms, etc (Yahoo!, 2000)– gaming financial accounts (PayPal, 2001)– robot spamming (MailBlocks, SpamArrest 2002)– In the last few months: Overture, Chinese website, HotMail, CD-rebate, TicketMaster, MailFrontier, Qurb, Madonnarama, …
…have you seen others?
On the horizon: – ballot stuffing, password guessing, denial-of-service attacks– `blunt force’ attacks (e.g. UT Austin break-in, Mar ’03)– …many others
Similar problems w/ scrapers; also, likely on Intranets.D. P. Baron, “eBay and Database Protection,” Case No. P-33, Case Writing Office, Stanford Graduate School of Business, Stanford Univ., 2001.
DIAR, Madison, WI – June 21, 2003 (HSB) 13
The Known Limits ofThe Known Limits ofImage Understanding TechnologyImage Understanding Technology
There remains a large gap in ability
between human and machine vision systems,
even when reading printed text
Performance of OCR machines has been systematically studied: 7 year olds can consistently do better!
This ability gap has been mapped quantitatively
S. Rice, G. Nagy, T. Nartker, OCR: An Illustrated Guide to the Frontier, Kluwer Academic Publishers: 1999.
DIAR, Madison, WI – June 21, 2003 (HSB) 14
Image Degradation ModelingImage Degradation Modeling
Effects of printing & imaging:
We can generate challenging
images pseudorandomly
H. Baird, “Document Image Defect Models,” in H. Baird, H. Bunke, & K. Yamamoto (Eds.), Structured Document Image Analysis, Springer-Verlag: New York, 1992.
blur
thrs
sen
s
thrs x blur
DIAR, Madison, WI – June 21, 2003 (HSB) 15
Machine Accuracy is a SmoothMachine Accuracy is a SmoothMonotonic Function of ParametersMonotonic Function of Parameters
T. K. Ho & H. S. Baird, “Large Scale Simulation Studies in Image Pattern Recognition,” IEEE Trans. on PAMI, Vol. 19, No. 10, p. 1067-1079, October 1997.
DIAR, Madison, WI – June 21, 2003 (HSB) 16
Can You ReadCan You Read These Degraded Images? These Degraded Images?
Of course you can …. but OCR machines cannot!
DIAR, Madison, WI – June 21, 2003 (HSB) 17
Experiments by PARC & UCB-CSExperiments by PARC & UCB-CS Pick words at random:
– 70 words commonly used on the Web
– w/out ascenders or descenders (cf. Spitz)
Vary physics-based image degradation parameters:
blur, threshold, x-scale -- within certain ranges
Pick fonts at random from a large set:
Times Roman (TR), Times Italic (TI), Palatino Roman (PR), Palatino Italic (PI), Courier Roman (CR), Courier Oblique (CO), etc
Test legibility on:– ten human volunteers (UC Berkeley CS Dept grad students)
– three OCR machines:
Expervision TR (E), ABBYY FineReader (A), IRIS Reader (I)
DIAR, Madison, WI – June 21, 2003 (HSB) 18
Results:Results: OCR Accuracy, by machine OCR Accuracy, by machine
00.10.20.30.40.50.60.70.80.9
1
fractionof wordscorrect
Expervis'n ABBYY IRIS
OCR machine
Times R
Times I
Courier O
Palatino R
Palatino I
total
Each machine has its peculiar blind spots
DIAR, Madison, WI – June 21, 2003 (HSB) 19
OCR Accuracy:OCR Accuracy: varying blur & threshold varying blur & threshold
0
0.2
0.4
0.6
0.8
1
fraction correct
0.02 0.04 0.06 0.08
threshhold
E blur=0.0E blur=0.4
E blur=0.8A blur=0.0A blur=0.4
A blur=0.8
The machines share some blind spots
DIAR, Madison, WI – June 21, 2003 (HSB) 20
PessimalPrint:PessimalPrint: exploiting image degradations exploiting image degradations
Three OCR machines fail when: OCR outputs
– blur = 0.0
& threshold 0.02 - 0.08
– threshold = 0.02
& any value of blur
~~~.I~~~
~~i1~~
N/A
N/A
N/A
~~I~~
A. Coates, H. Baird, R. Fateman, “Pessimal Print: A Reverse Turing Test,” Proc. 6th IAPR Int’l Conf. On Doc. Anal. & Recogn. (ICDAR’01), Seattle, WA, Sep 10-13, 2001.
… but people find all these easy to read
DIAR, Madison, WI – June 21, 2003 (HSB) 21
High Time for a Workshop!High Time for a Workshop!
Manuel Blum proposes it, rounds up some key speakers
Henry Baird offers PARC as venue; Kris Popat helps run it
Goals:Invite all known principals: theory, systems, engineers, users
Describe the state of the art
Plan next steps for the field
Organization:– ~30 attendees
– abstracts only, 1-5 pages, no refereeing, no archival publication
– 100% participation: everyone gives a (short) talk
– “mixing it up”: panel & working group discussions
– 2-1/2 days, lots of breaks for informal socializing
– plenary talk by John McCarthy ‘Father of AI’
DIAR, Madison, WI – June 21, 2003 (HSB) 22
1st NSF Int’l Workshop on1st NSF Int’l Workshop onHuman Interactive ProofsHuman Interactive Proofs PARC, Palo Alto, CA, January 9-11, 2002PARC, Palo Alto, CA, January 9-11, 2002
DIAR, Madison, WI – June 21, 2003 (HSB) 23
HIP’2002 ParticipantsHIP’2002 ParticipantsCMU - SCS, Aladdin Center
Manuel Blum, Lenore Blum, Luis von Ahn, John Langford, Guy Blelloch, Nick Hopper, Ke Yang, Brighten Godfrey, Bartosz Przydatek, Rachel Rue
PARC - SPIA/Security/TheoryHenry Baird, Kris Popat, Tom Breuel,
Prateek Sarkar, Tom Berson, Dirk Balfanz, David Goldberg
UCB - CS & SIMSRichard Fateman, Allison Coates,
Jitendra Malik, Doug Tygar, Alma Whitten, Rachna Dhamija, Monica Chew, Adrian Perrig, Dawn Song
RPIGeorge Nagy
StanfordJohn McCarthy
NSFRobert Sloan
AltavistaAndrei Broder
Yahoo!Udi Manber
Bell LabsDan Lopresti
IBM T.J. WatsonCharles Bennett
InterTrust Star LabsStuart Haber
City Univ. of Hong HongNancy Chan
Weizmann InstituteMoni Naor
RSA Security LaboratoriesAri Juels
Document Recognition Techs, IncLarry Spitz
DIAR, Madison, WI – June 21, 2003 (HSB) 24
Variations & GeneralizationsVariations & Generalizations CAPTCHA
Completely Automatic Public Turing test to tell Computers and Humans Apart
HUMANOIDText-based dialogue which an individual can use to
authenticate that he/she is himself/herself (‘naked in a glass bubble’)
PHONOIDIndividual authentication using spoken language
Human Interactive Proof (HIP)An automatically administered challenge/response protocolAn automatically administered challenge/response protocol
allowing a person to authenticate him/herself as belonging to allowing a person to authenticate him/herself as belonging to a certain group over a network without the burden of a certain group over a network without the burden of passwords,passwords,
biometrics, mechanical aids, or special training.biometrics, mechanical aids, or special training.
DIAR, Madison, WI – June 21, 2003 (HSB) 25
Highlights of HIP’2002Highlights of HIP’2002
Theory– some text-based CAPTCHAs are provably breakable
Ability Gaps– vision: gestalt, segmentation, noise immunity, style consistency
– speech: noise of many kinds, clutter (cocktail party effect)
– intelligence: puzzles, analogical reasoning, weak logic
– gestures, reflexes, common knowledge, …
Applications– subtle system-level vulnerabilties– aggressive arms race with shadowy enemies
http://www.parc.com/istl/groups/did/HIP2002
DIAR, Madison, WI – June 21, 2003 (HSB) 26
Funding & PartnershipsFunding & Partnerships
NSF– Robert Sloan, Dir, Theory of Computing Pgm– strongly supportive of this newborn field– encouraged grant proposals
Yahoo!– willing to run field trials– user acceptance laboratory– able to detect intrusion
DIAR, Madison, WI – June 21, 2003 (HSB) 27
DisciplinesDisciplines
Participating now: Cryptography Security Pattern Recognition Computer Vision Artificial Intelligence eCommerce
Needed: Cognitive Science Psychophysics (esp. of Reading) Biometrics Business, Law, … ….?
DIAR, Madison, WI – June 21, 2003 (HSB) 28
Weaknesses of ExistingWeaknesses of Existing Reading-Based CAPTCHAs Reading-Based CAPTCHAs
English lexicon is too predictable:– dictionaries are too small– only 1.2 bits of entropy per character (cf. Shannon)
Physics-based image degradations vulnerable to well-studied image restoration attacks, e.g.
Complex images irritate people
– even when they can read them– need user-tolerance experiments
DIAR, Madison, WI – June 21, 2003 (HSB) 29
Strengths ofStrengths of Human Reading Human Reading
Literature on the psychophysics of reading is relevant:
familiarity helps, e.g. English words optimal word-image size (subtended angle) is known (0.3-2 degrees) optimal contrast conditions known other factors measured for the best performance:
to achieve and sustain “critical reading speed”
BUT gives no answer to: where’s the optimal comfort zone?
G. E. Legge, D. G. Pelli, G. S. Rubin, & M. M. Schleske, “Psychophysics of Reading: I. normal vision,” Vision Research 25(2), 1985.
A. J. Grainger & J. Segui, “Neighborhood Frequency Effects in Visual Word Recognition,’ Perception & Psychophysics 47, 1990..
DIAR, Madison, WI – June 21, 2003 (HSB) 30
Designing a Stronger CAPTCHA:Designing a Stronger CAPTCHA: BaffleTextBaffleText principles principles
Nonsense words.– generate ‘pronounceable’ – not ‘spellable’ – words using a variable-length character n-gram Markov model– they look familiar, but aren’t in any lexicon, e.g.
ablithan wouquire quasis
Gestalt perception.– force inference of a whole word-image from fragmentary or occluded characters, e.g.
– using a single familiar typeface also helps
M. Chew & H. S. Baird, “BaffleText: A Human Interactive Proof,” Proc., SPIE/IS&T Conf. on Document Recognition & Retrieval X, Santa Clara, CA, January 23-24, 2003.
DIAR, Madison, WI – June 21, 2003 (HSB) 31
Mask DegradationsMask Degradations
Parameters of pseudorandom mask generator:– shape type: square, circle, ellipse, mixed– density: black-area / whole-area– range of radii of shapes
DIAR, Madison, WI – June 21, 2003 (HSB) 32
BaffleTextBaffleText Experiments at PARC Experiments at PARC
Goal: map the margins of accurate & comfortable
human reading on this family of images Metrics:
– objective difficulty: accuracy– subjective difficulty: rating– response time– exit survey: how tolerable overall
Participation:– 41 individual sessions– >1200 challenge/response trials– 18 exit surveys
DIAR, Madison, WI – June 21, 2003 (HSB) 33
BaffleTextBaffleText challenge webpage challenge webpage
DIAR, Madison, WI – June 21, 2003 (HSB) 34
BaffleTextBaffleText user ratings user ratings
DIAR, Madison, WI – June 21, 2003 (HSB) 35
User AcceptanceUser Acceptance
% Subjects willing to solve a BaffleText…
17% every time they send email
39% … if it cut spam by 10x
89% every time they register for an e-commerce site
94% … if it led to more trustworthy recommendations
100% every time they register for an email account
Out of 18 responses to the exit survey.
DIAR, Madison, WI – June 21, 2003 (HSB) 36
Subjective difficultySubjective difficulty tracks objective difficulty tracks objective difficulty
DIAR, Madison, WI – June 21, 2003 (HSB) 37
How to engineer How to engineer BaffleTextBaffleText
When we generate a challenge,– need to estimate its difficulty– throw away if too easy or too hard
Apply an idea from the psychophysics of reading:– image “complexity” metric: how hard to read
– simple to compute: perimeter** / black-area
DIAR, Madison, WI – June 21, 2003 (HSB) 38
Image complexityImage complexity predicts objective difficulty predicts objective difficulty
DIAR, Madison, WI – June 21, 2003 (HSB) 39
Image complexityImage complexity predicts subjective difficulty predicts subjective difficulty
DIAR, Madison, WI – June 21, 2003 (HSB) 40
Engineering guidelinesEngineering guidelines
For high performance, image complexity
should fall in the range 50-100; e.g.
Within this regime, BaffleText performs well:– 100% human subjects willing to try to read it– 89% accuracy by humans– 0% accuracy by commercial OCR– 3.3 difficulty rating, out of 10 (on average)– 8.7 seconds / trial on average
50 100
DIAR, Madison, WI – June 21, 2003 (HSB) 41
The latest seriousThe latest serious (known or published) attack… (known or published) attack…
Greg Mori & Jitendra Malik (UCB-CS)– Generalized Shape Context CV method– requires known lexicon – else, fails completely– expects known font (or fonts) – else, does worse
Results of Mori-Malik attacks (Dec 2002) given perfect foreknowledge of both lexicon and font:
CAPTCHA Attack success rate
EZ-GIMPY
Yahoo! + CMU83%
PessimalPrint
PARC + UCB40%
BaffleText PARC + UCB
25%
G. Mori & J. Malik, “Recognizing Objects in Adversarial Clutter,” submitted to CVPR’03, Madison, WI, June 16-22, 2003.
DIAR, Madison, WI – June 21, 2003 (HSB) 42
BaffleText:BaffleText: the strongest known CAPTCHA? the strongest known CAPTCHA?
Resists many known algorithmic attacks:– physics-based image restoration
– recognizing into a lexicon
– known-typeface targeting
– segmenting then recognizing
Exploits hard-to-automate human cognition powers:– Gestalt perception
– “semi-linguistic” familiarity
– within-typeface “style consistency”
DIAR, Madison, WI – June 21, 2003 (HSB) 43
Recent Microsoft CAPTCHARecent Microsoft CAPTCHA
• Random strings, local space-warping; plus meaningless curving strokes, both black (overlaid) and white (erasing)
• Fielded Dec 2002 on Passport (HotMail, etc)• Immediate reduction in new Hotmail accounts, with
virtually no user complaints
P. Y. Simard, R. Szeliski, J. Benaloh, J. Couvreur, I. Calinov, “Using Character Recognition and Segmentation to Tell Computer from Humans,” Proc., Int’l Conf. on Document Analysis & Recognition, Edinburgh, Scotland, August, 2003 [to appear].
DIAR, Madison, WI – June 21, 2003 (HSB) 44
PARC’s Leadership in R&D onPARC’s Leadership in R&D on Reading-based CAPTCHAs Reading-based CAPTCHAs
First refereed article on CAPTCHAs: A. L. Coates, H. S. Baird, R. Fateman, “Pessimal Print: a Reverse Turing Test,” Proc., 6th
IAPR Int’l Conf. On Document Analysis & Recognition, Seattle, WA, Sept. 10-13, 2001.
First professional HIP event, organized by PARC:
1st NSF Int’l Workshop on HIPs, Jan. 9-11, 2002, PARC, Palo Alto, CA.
First to ‘play both offense & defense’:– builds high-performance OCR systems; attacks CAPTCHAs– builds strong CAPTCHAs
First to validate using human-factors research:– human-subject trials measuring both accuracy & tolerance– PARC’s interdisciplinary tradition: social + computer sciences
DIAR, Madison, WI – June 21, 2003 (HSB) 45
The Arms Race The Arms Race
When will serious technical attacks be launched?
– ‘spam kings’ make $$ millions
– two spam-blocking e-commerce firms now use CAPTCHAs
How long can a CAPTCHA withstand attack?
– especially if its algorithms are published or guessed
Strategy: keep a pipeline of defenses in reserve:
– continuing partnership between R&D & users
DIAR, Madison, WI – June 21, 2003 (HSB) 46
Lots of Open Research QuestionsLots of Open Research Questions
What are the most intractable obstacles to machine vision?
segmentation, occlusion, degradations, …?
Under what conditions is human reading most robust?
linguistic & semantic context, Gestalt, style consistency…?
Where are ‘ability gaps’ located?
quantitatively, not just qualitatively
How to generate challenges strictly within ability gaps?
fully automatically
an indefinitely long sequence of distinct challenges
DIAR, Madison, WI – June 21, 2003 (HSB) 47
HIP Research CommunityHIP Research Community
PARC CAPTCHA website
www.parc.com/istl/projects/captcha
HIP’2002 Workshop
www.parc.com/istl/groups/did/HIP2002
HIP Website at Aladdin Center, CMU-SCS www.captcha.net
Volunteers for a PARC CAPTCHA usability test?
A 2nd HIP Workshop soon?
DIAR, Madison, WI – June 21, 2003 (HSB) 48
Alan Turing might haveAlan Turing might have enjoyed the irony … enjoyed the irony …
A technical problem – machine reading –
which he thought would be easy,
has resisted attack for 50 years, and
now allows the first widespread
practical use of variants of
his test for artificial intelligence.
DIAR, Madison, WI – June 21, 2003 (HSB) 49
ContactContact
Henry S. Baird
www.parc.com/baird