JUICER: Data-Efficient Imitation Learning for Robotic Assembly

arXiv:2404.03729v1 Announce Type: cross
Abstract: While learning from demonstrations is powerful for acquiring visuomotor policies, high-performance imitation without large demonstration datasets remains challenging for tasks requiring precise, long-horizon manipulation. This paper proposes a pipeline for improving imitation learning performance with a small human demonstration budget. We apply our approach to assembly tasks that require precisely grasping, reorienting, and inserting multiple parts over long horizons and multiple task phases. Our pipeline combines expressive policy architectures and various techniques for dataset expansion …

abstract apply arxiv assembly budget cs.lg cs.ro data datasets horizon human imitation learning improving manipulation paper performance pipeline policies robotic small tasks type