双臂家务操作：从 1500 小时演示扩展到策略内纠正
Original title: Scaling Bimanual Household Manipulation from 1,500 hours of Demonstrations to On-Policy Corrections
ResearchAI 88
Source: arXiv cs.ROPublish time unverified
arXiv:2609.03591v1 Announce Type: new Abstract: Learning generalist policies for robust bimanual manipulation is bottlenecked by the scarcity of high quality large scale human demonstration data. In this work, we release 1,500 hours of diverse bimanual manipulation demonstrations covering everyday household tasks, and use this comprehensive corpus to train XR-2, a powerful vision-language-action (VLA) model.