Reproduce “From Hitter to Player: Separating Motor Execution and Strategy in Competitive Humanoid Table Tennis” 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
Fanrong Dong, Ziyi Sun, Ziqi Han, Junhan Sun, Lei Ye, Zetong Jing, Hao Zhao, Shaoxun Liu and 1 more
The authors present a robot system that separates target selection from stroke execution through a shared landing-target command. The motor policy learns from prescribed balls and closed-loop self-play, then remains fixed while a strategic policy selects among nine regions. In simulation, the learned selector wins 73.1% of points against Away, under matched execution; this is not evidence of human-level play or generalization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.