Embodiment-Aware Generalist Specialist Distillation for Unified Humanoid Whole-Body Control

2026-02-03

AI Summary

Existing reinforcement learning policies struggle to generalize across heterogeneous humanoid robots and support diverse whole-body motions. To address this challenge, this work proposes the EAGLE framework, which introduces a novel embodied perception–based general-to-specialized policy distillation mechanism. By iteratively distilling a universal policy with robot-specific specialized policies, EAGLE achieves cross-morphology policy generalization and multi-skill integration without requiring per-robot hyperparameter tuning. The framework enables unified control of various humanoid robots to execute complex maneuvers such as squatting and leaning. Experiments on five simulated and four real-world robots demonstrate that EAGLE significantly outperforms existing methods in motion tracking accuracy and robustness, advancing the state of the art in scalable, cluster-level humanoid robot control.

Technology Category

• Intelligent Robots: Embodied AI
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Humanoid Whole-Body Controllers trained with reinforcement learning (RL) have recently achieved remarkable performance, yet many target a single robot embodiment. Variations in dynamics, degrees of freedom (DoFs), and kinematic topology still hinder a single policy from commanding diverse humanoids. Moreover, obtaining a generalist policy that not only transfers across embodiments but also supports richer behaviors-beyond simple walking to squatting, leaning-remains especially challenging. In this work, we tackle these obstacles by introducing EAGLE, an iterative generalist-specialist distillation framework that produces a single unified policy that controls multiple heterogeneous humanoids without per-robot reward tuning. During each cycle, embodiment-specific specialists are forked from the current generalist, refined on their respective robots, and new skills are distilled back into the generalist by training on the pooled embodiment set. Repeating this loop until performance convergence produces a robust Whole-Body Controller validated on robots such as Unitree H1, G1, and Fourier N1. We conducted experiments on five different robots in simulation and four in real-world settings. Through quantitative evaluations, EAGLE achieves high tracking accuracy and robustness compared to other methods, marking a step toward scalable, fleet-level humanoid control. See more details at https://eagle-wbc.github.io/

Problem

• humanoid whole-body control
• embodiment generalization
• multi-robot policy
• reinforcement learning
• cross-embodiment transfer

Innovation

• embodiment-aware
• generalist-specialist distillation
• whole-body control
• humanoid robotics
• policy transfer