Prompt, Plan, Perform: LLM-based Humanoid Control via Quantized Imitation Learning

TL;DR

This paper introduces a novel humanoid robot control framework combining adversarial imitation learning, large language models as planners, and vector quantization to enable zero-shot task execution with a single policy.

Contribution

It presents the first framework integrating LLMs as strategic planners with a unified policy for humanoid control, enhancing adaptability and reusability.

Findings

Efficient zero-shot task execution demonstrated in experiments.

Single policy network effectively controls complex humanoid motions.

Incorporation of vector quantization improves action generation for unseen commands.

Abstract

In recent years, reinforcement learning and imitation learning have shown great potential for controlling humanoid robots' motion. However, these methods typically create simulation environments and rewards for specific tasks, resulting in the requirements of multiple policies and limited capabilities for tackling complex and unknown tasks. To overcome these issues, we present a novel approach that combines adversarial imitation learning with large language models (LLMs). This innovative method enables the agent to learn reusable skills with a single policy and solve zero-shot tasks under the guidance of LLMs. In particular, we utilize the LLM as a strategic planner for applying previously learned skills to novel tasks through the comprehension of task-specific prompts. This empowers the robot to perform the specified actions in a sequence. To improve our model, we incorporate codebook-based vector quantization, allowing the agent to generate suitable actions in response to unseen textual commands from LLMs. Furthermore, we design general reward functions that consider the distinct motion features of humanoid robots, ensuring the agent imitates the motion data while maintaining goal orientation without additional guiding direction approaches or policies. To the best of our knowledge, this is the first framework that controls humanoid robots using a single learning policy network and LLM as a planner. Extensive experiments demonstrate that our method exhibits efficient and adaptive ability in complicated motion tasks.