Presentation

<
Hand-Object Interaction Controller (HOIC): Deep Reinforcement Learning for Reconstructing Interactions With Physics
SessionMotion Capture
DescriptionThis paper introduces a physics-based hand-object interaction reconstruction system by leveraging imitation learning. A novel object compensation control technique is proposed to upgrade the simple point contact model to a more physical-plausible surface contact model, which improves the training stability and reconstruction quality.
Authors
Haoyu Hu
School of Software, Tsinghua University
Xin-yu Yi
School of Software, Tsinghua University
Zhe Cao
Google
Jun-Hai Yong
School of Software, Tsinghua University
Feng Xu
School of Software, Tsinghua University
Event Type
Technical Paper
TimeWednesday, 31 July 20249:40am - 9:50am MDT
LocationMile High 4
ACM Digital Library Journal Papers' PDFs
Conference Papers' PDFs
Session Time & Location
Sunday, 28 July 20246:00pm - 8:45pm MDTBluebird Ballroom
Wednesday, 31 July 20249:00am - 10:30am MDTMile High 4
Interest Areas
Research & Education
Recordings
Livestreamed
Recorded
Keywords
Animation
Modeling
Registration Categories
Full Conference
Full Conference Supporter
Virtual Access
Exhibitor Full Conference
Wednesday
Next PresentationNext PresentationPhysics-informed Learning of Characteristic Trajectories for Smoke Reconstruction