Reproduce “ReContact: Reconstruct and Retarget Human Contacts from Video for Contact-Rich Humanoid Manipulation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhongkai Gu, Yang Liu, Wei Zhu, Mitsuhiro Hayashibe
ReContact represents manipulation as phased contact plans that specify robot–object pairs, object-surface anchors, pressing directions, and sticking or sliding. The plan guides object reconstruction, robot retargeting, and residual control. On a shared simulated thigh-supported regrasp reference, it achieves 83.52 ± 5.20% prescribed two-palm bottom-hold success, versus 0.16% for the best adapted baseline. This result concerns that task and reference, not general performance.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.