C-3PO: Cyclic-Three-Phase Optimization for Human-Robot Motion Retargeting based on Reinforcement Learning

TL;DR

This paper introduces C-3PO, a cyclic three-phase deep reinforcement learning method for human-robot motion retargeting across different robot types, improving policy learning through a human-in-the-loop framework.

Contribution

The paper presents a novel cyclic three-phase optimization approach that enables kinematic-independent motion retargeting using deep reinforcement learning and a human-in-the-loop framework.

Findings

Successfully retargeted motion between human skeleton and multiple robots

Improved motion retargeting policy through human-in-the-loop feedback

Demonstrated effectiveness on robots like NAO, Pepper, Baxter, and C-3PO

Abstract

Motion retargeting between heterogeneous polymorphs with different sizes and kinematic configurations requires a comprehensive knowledge of (inverse) kinematics. Moreover, it is non-trivial to provide a kinematic independent general solution. In this study, we developed a cyclic three-phase optimization method based on deep reinforcement learning for human-robot motion retargeting. The motion retargeting learning is performed using refined data in a latent space by the cyclic and filtering paths of our method. In addition, the human-in-the-loop based three-phase approach provides a framework for the improvement of the motion retargeting policy by both quantitative and qualitative manners. Using the proposed C-3PO method, we were successfully able to learn the motion retargeting skill between the human skeleton and motion of the multiple robots such as NAO, Pepper, Baxter and C-3PO.