Physics-Based Motion Tracking of Contact-Rich Interacting Characters
2026-04-09 • Graphics
Graphics
AI summaryⓘ
The authors study how to make computer-generated characters move like people, especially when they interact closely with others or objects. Existing methods usually focus on just one character or simple interactions and have trouble when things touch or push each other a lot. To fix this, the authors use a setup where multiple expert models learn different skills, and the system decides which expert to use without needing manual instructions. Their approach results in smoother and more stable motions during complex interactions and also trains the models more efficiently.
motion trackingphysics-based motion synthesiscontact-rich interactionsneural networksprogressive neural networkmulti-expert modelmotion stabilitymodel traininginteraction forcescharacter animation
Authors
Xiaotang Zhang, Ziyi Chang, Qianhui Men, Hubert P. H. Shum
Abstract
Motion tracking has been an important technique for imitating human-like movement from large-scale datasets in physics-based motion synthesis. However, existing approaches focus on tracking either single character or a particular type of interaction, limiting their ability to handle contact-rich interactions. Extending single-character tracking approaches suffers from the instability due to the challenge of forces transferred through contacts. Contact-rich interactions requires levels of control, which places much greater demands on model capacity. To this end, we propose a robust tracking method based on progressive neural network (PNN) where multiple experts are specialized in learning skills of various difficulties. Our method learns to assign training samples to experts automatically without requiring manually scheduling. Both qualitative and quantitative results show that our method delivers more stable motion tracking in densely interactive movements while enabling more efficient model training.