Reproduce “Learning Object-Centric Motion Priors from Human for Robotic Dexterous 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
Zhengdong Hong, Guofeng Zhang
The authors present a method for guiding dexterous robot manipulation with predicted object motion from human interaction data. A transformer predicts future hand–object states, and predicted object states guide reinforcement learning in simulation, alongside task-success and contact rewards. On real-robot YCB grasping, it averages 0.77 success versus 0.55 for AdaDexGrasp across three categories; evidence covers three tasks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.