From One Hand to Multiple Hands: Imitation Learning for Dexterous Manipulation From Single-Camera Teleoperation

We propose to perform imitation learning for dexterous manipulation with multi-finger robot hand from human demonstrations, and transfer the policy real hand. introduce a novel single-camera teleoperation system collect 3D demonstrations efficiently only an iPad computer. One key contribution of our is that we construct customized each user in simulator, which manipulator resembling same structure operator's It provides intuitive interface avoid unstable human-robot retargeting data collection, leading large-scale high quality data. Once collected, trajectories can be converted different specified hands (models are manufactured) generate training demonstrations. With using data, show large improvement over baselines multiple complex tasks. Importantly, learned significantly more robust when transferring robot.