Reproduce “ARMFUL: Learning Whole-Arm Grasp Configurations for Single-Arm Pickup of Large Objects” 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
Xinyi Yu, Yunshuang Li, Daniel Seita, Gaurav S. Sukhatme
The authors introduce ARMFUL to propose whole-arm grasps for tabletop pickup. They build WAG by generating and stress-testing configurations in simulation, then train a model to select grasps and plan approaches; squeeze-and-lift motion remains fixed. On held-out objects, lift success is 69.86% versus 58.00% for nearest-neighbor retrieval; a Franka lifts IKEA objects absent from WAG in 70 of 100 trials.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.