Reproduce “Ctrl-F: Force as an Interface for Humanoid Loco-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
Mengjie Zhao, Yuxin Chen, Yiyang Shao, Chuye Hong, Jiakang Chen, Koushil Sreenath, Masayoshi Tomizuka
Ctrl-F pairs a humanoid tracker, which follows reference motion while executing three-dimensional force commands at designated body parts, with a planner that selects commands from object state and robot proprioception. The planner needs no force measurements or object mass at deployment, but requires object-state estimates. Across five hand tasks, completion stays above 90% from 0.5× to 1.5× nominal mass; force-blind baselines reach zero at 1.5×.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.