motion modeling for online locomotion synthesis

26
Motion Modeling for Online Locomotion Synthesis Taesoo Kwon and Sung Yong Shin KAIST

Upload: nusa

Post on 12-Jan-2016

36 views

Category:

Documents


0 download

DESCRIPTION

Motion Modeling for Online Locomotion Synthesis. Taesoo Kwon and Sung Yong Shin KAIST. Outline. Motivation Related work Overview Motion analysis Motion synthesis Conclusions Future Work. Motivation. Real-time locomotion synthesis Motion rearrangement : realism - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Motion Modeling for Online Locomotion Synthesis

Motion Modeling for Online Locomotion

Synthesis

Taesoo Kwon and Sung Yong Shin

KAIST

Page 2: Motion Modeling for Online Locomotion Synthesis

Outline

• Motivation

• Related work

• Overview

• Motion analysis

• Motion synthesis

• Conclusions

• Future Work

Page 3: Motion Modeling for Online Locomotion Synthesis

Motivation

• Real-time locomotion synthesis

• Motion rearrangement : realism

• Motion blending : efficiency and controllability

• Hybrid approach– Locomotive motion generation [PSS02, PSS04]– Rhythmic motion synthesis [KPS03]

• Premise: motion labeling

Page 4: Motion Modeling for Online Locomotion Synthesis

Related Work

• Motion Segmentation [Bindiganavale & Badler, 1998;Fod et al., 2002;

Kim et al., 2003]

• Motion Classification [Arikan et al., 2003;Kovar & Gleicher, 2004;Forbes & Fiu

me 2005;Mueller & Roeder 2005]

• Motion Labeling for blending [Kim et al., 2003]

Page 5: Motion Modeling for Online Locomotion Synthesis

Overview

motion specifications

desiredmotion

example motions

motion analysis

hierarchical motion transition graph

motion synthesis

Page 6: Motion Modeling for Online Locomotion Synthesis

Motion Analysis

• Issues– Motion segmentation

– Motion classification

– Graph construction

• Biomechanical observations– [Per92,Win90]

Page 7: Motion Modeling for Online Locomotion Synthesis

Biomechanical Observations• Center of mass trajectory

right foot left foot

walk runtransition

COMy

Page 8: Motion Modeling for Online Locomotion Synthesis

Motion segmentation

• Criteria for motion segmentation– Simple enough for intuitive parameterization

– Long enough to contain motion semantics

– An important motion feature should not be split

Split at every COM peak

Page 9: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• String encoding–

• Pros– avoid troublesome time-warping

– more robust than numerical computation

M:f

Page 10: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• Footstep patterns

(a) S (b) R (c) L (d) D (e) F FDLRS , , , ,

Page 11: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• String Encoding (ideal case)

Page 12: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• String Encoding (ideal case)

R D L

Page 13: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• String Encoding (ideal case)

F R F

Page 14: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• String Encoding (ideal case)

R D L F

Page 15: Motion Modeling for Online Locomotion Synthesis

Motion Classification

• String Encoding (ideal case)

Page 16: Motion Modeling for Online Locomotion Synthesis

Refinement

• False peak– Concatenate two motion segments

• Missing peak– Divide a motion segment into two

Page 17: Motion Modeling for Online Locomotion Synthesis

Graph Construction

Page 18: Motion Modeling for Online Locomotion Synthesis

Graph Construction

mmmmmmP avftt ,,,,

Page 19: Motion Modeling for Online Locomotion Synthesis

Motion Analysis Results

• O(n) – 2Ghz PC (AMD 64, 2GB memory)

– For 7.4 min locomotion, about 10 seconds

• Movie

Page 20: Motion Modeling for Online Locomotion Synthesis

Motion Synthesis

LDR RDL LDRF… …

Page 21: Motion Modeling for Online Locomotion Synthesis

Motion Synthesis

• Motion specification

• Motion parameter

Page 22: Motion Modeling for Online Locomotion Synthesis

Motion Sythesis

• How to calculate – Two half cycles in cyclic motion

• Regression analysis on

m

RL mm

RL mm

Page 23: Motion Modeling for Online Locomotion Synthesis

Motion Synthesis

• Motion blending : [PSS04][KG03][ACP02]

• Motion stitching : [GSKJ03]

• Motion retargeting : [SLSG01][KGS02]

Page 24: Motion Modeling for Online Locomotion Synthesis

Motion Synthesis Result

• 1000+ frames per second

• Movie– Path following

– Online synthesis

Page 25: Motion Modeling for Online Locomotion Synthesis

Conclusion

• Motion labeling based on string encodings

• Hierarchical motion transition graph

Page 26: Motion Modeling for Online Locomotion Synthesis

Future work

• Footstep-driven motions such as dancing and boxing