Reproduce “RoboFolDeX: A Physical-World Benchmark for Long-Horizon Robotic Manipulation of Deformable 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
Chenhuan Liu, Yi Xu, Feng Wu, Hanyang Wang, Wenxiao Kuai, Weihao Ding, Shan Wang, Yang Liu and 2 more
RoboFolDeX presents a real-robot benchmark and external evaluation protocol for complete garment-folding episodes. It organizes physical data across recovery, task, scene, and embodiment reuse tracks, with more than 2,000 hours across 20+ tasks and 10+ robotic embodiments. In a preliminary comparison, adding recovery trajectories was associated with average success rising from 80.75% to 95.00% for the tested policy and protocol.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.