Hierarchical World Models as Visual Whole-Body Humanoid Controllers

Autor:	Hansen, Nicklas, S V, Jyothir, Sobal, Vlad, LeCun, Yann, Wang, Xiaolong, Su, Hao
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning Computer Science - Computer Vision and Pattern Recognition Computer Science - Robotics
Druh dokumentu:	Working Paper
Popis:	Whole-body control for humanoids is challenging due to the high-dimensional nature of the problem, coupled with the inherent instability of a bipedal morphology. Learning from visual observations further exacerbates this difficulty. In this work, we explore highly data-driven approaches to visual whole-body humanoid control based on reinforcement learning, without any simplifying assumptions, reward design, or skill primitives. Specifically, we propose a hierarchical world model in which a high-level agent generates commands based on visual observations for a low-level agent to execute, both of which are trained with rewards. Our approach produces highly performant control policies in 8 tasks with a simulated 56-DoF humanoid, while synthesizing motions that are broadly preferred by humans. Code and videos: https://nicklashansen.com/rlpuppeteer Comment: Code and videos at https://nicklashansen.com/rlpuppeteer
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2405.18418 Zobrazit plný text záznamu View this record from Arxiv