H-Zero: Cross-Humanoid Locomotion Pretraining Enables Few-shot Novel Embodiment Transfer

TL;DR

H-Zero is a pretraining framework that enables zero-shot and few-shot transfer of humanoid locomotion controllers to new robot embodiments with minimal fine-tuning, reducing development time and effort.

Contribution

The paper introduces H-Zero, a novel cross-humanoid pretraining pipeline that generalizes locomotion policies across different robot designs, facilitating rapid adaptation.

Findings

Pretrained policy achieves up to 81% of full episode duration on unseen robots.

Enables few-shot transfer within 30 minutes of fine-tuning.

Reduces need for extensive tuning for each new humanoid robot.

Abstract

The rapid advancement of humanoid robotics has intensified the need for robust and adaptable controllers to enable stable and efficient locomotion across diverse platforms. However, developing such controllers remains a significant challenge because existing solutions are tailored to specific robot designs, requiring extensive tuning of reward functions, physical parameters, and training hyperparameters for each embodiment. To address this challenge, we introduce H-Zero, a cross-humanoid locomotion pretraining pipeline that learns a generalizable humanoid base policy. We show that pretraining on a limited set of embodiments enables zero-shot and few-shot transfer to novel humanoid robots with minimal fine-tuning. Evaluations show that the pretrained policy maintains up to 81% of the full episode duration on unseen robots in simulation while enabling few-shot transfer to unseen humanoids and upright quadrupeds within 30 minutes of fine-tuning.