Reproduce “DexPort: A Hand-Held Skill Interface for Dexterous Hand Data Collection” 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
Amber Xie, Jiawei Fu, Xiaofeng Guo, Carmelo Sferrazza, Guanya Shi, Haozhi Qi
DexPort enables portable demonstrations for a 20-DoF robot hand by letting operators guide its wrist and control coordinated finger skills with the same hand used at deployment. Policies learn from commanded joint targets, not collection skills. Across three success-scored tasks, command-supervised policies average 64% final success versus 11% for measured-state labels; this result is limited to the tested setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.