CSI:Integrating Controllable Motion Skills from Demonstrations
School of Control Science and Engineering, Shandong University
Abstract
The expanding applications of legged robots require their mastery of versatile motion skills. Correspondingly, researchers must address the challenge of integrating multiple diverse motion skills into controllers. While existing reinforcement learning~(RL)-based approaches have achieved notable success in multi-skill integration for legged robots, these methods often require intricate reward engineering or are restricted to integrating a predefined set of motion skills constrained by specific task objectives, resulting in limited flexibility. In this work, we introduce a flexible multi-skill integration framework named Controllable Skills Integration~(CSI). CSI enables the integration of a diverse set of motion skills with varying styles into a single policy without the need for complex reward tuning. Furthermore, in a hierarchical control manner, the trained low-level policy can be coupled with a high-level Natural Language Inference~(NLI) module to enable preliminary language-directed skill control. Our experiments demonstrate that CSI can flexibly integrate a diverse array of motion skills more comprehensively and facilitate the transitions between different skills. Additionally, CSI exhibits good scalability as the number of motion skills to be integrated increases significantly.
Overview
Framework
Description
Overview diagram of CSI. Through retargeting and skill labeling, a set of reference motion clips with corresponding labels can be obtained. During training, sampled motion skill labels $c_i$ are mapped to latent vectors $z$ through an embedding network, and the policy generates corresponding motion skills based on $z$. The discriminator is responsible for indirectly regulating the motions generated by the policy in a way that provides style rewards. After the training stage, a controller with integrated multiple motion skills is available. These integrated skills can be controlled directly through user commands or externally via a high-level pre-trained NLI module for language-directed skill control.
Comparison with Other Methods
Mode Collapse
Desmonstrations on Various Platforms
Visualization of Different Tasks
H-Locomotion
H-Multiwalk
Q-Locomotion
H-Interaction