The paper presents a study on object grasping in robotics using bilateral control-based imitation learning and a cross-structure hand. This technique allows robots to mimic human force skills and adapt to their environment without the need for programming. However, hardware limitations have restricted its performance. The study demonstrates that integrating this learning method with a cross-structure hand is effective for grasping various objects and using tools. This research is significant in the field of robotics, particularly in tasks that require precise force adjustments.
Publication date: 20 Nov 2023
Project Page: DOI: see top of this page
Paper: https://arxiv.org/pdf/2311.09555
