ADAPT: Agile Diffusion Action Priors for Robust and Steerable Online Text-Driven Humanoid Control

2026-08-31

AI Summary

ADAPT通过端到端闭环控制框架和基于扩散的动作先验学习，解决了文本驱动的人形机器人实时、自然且平衡的全身控制问题。

Abstract

We present ADAPT, an end-to-end framework for interactive, text-conditioned humanoid whole-body control. Unlike dominant text-to-motion pipelines that generate kinematic motions for a separate tracker, ADAPT solves language control with an end-to-end closed-loop control framework, where the robot must continuously respond to changing commands while maintaining balance, natural motion, and smooth transitions. ADAPT learns a diffusion-based action prior from text-labeled humanoid state-action trajectories, enabling diverse motion skills to be directly executed from language commands. To improve long-horizon robustness and smooth prompt switching, we train a lightweight residual reinforcement learning policy on top of the frozen diffusion controller. We further show that the same diffusion policy can be reused as a steerable text-conditioned motion prior for downstream task adaptation. Experiments demonstrate robust language-grounded skill execution, smooth interactive transitions, and style-preserving downstream control.

Problem

• text-conditioned
• humanoid control
• end-to-end
• closed-loop
• natural motion

Innovation

• end-to-end closed-loop control
• diffusion-based action prior
• residual reinforcement learning
• steerable text-conditioned motion prior