[论文] Perceptive Humanoid Motion from Simulation

论文概要

研究领域: CV 作者: Siyuan Hu, Abulikemu Abuduweili, Yue Meng, Shipeng Liu, Quentin Leboutet, Chen Chen, Matthias Müller, Vladlen Koltun 发布时间: 2026-11-04 arXiv: 2611.06908

中文摘要

人形机器人有望自动化广泛的物理任务。然而，通向通用人形自主性的进展受制于数据瓶颈：将像素和自由形式语言映射到低层关节指令需要成对监督数据，而目前不存在足够大规模的此类语料。一些方法依赖人类遥操作和可穿戴捕获设备，产生一小时数据至少需一小时人力。我们提出 HALO——由跨千个程序化生成家庭合成视觉-语言-运动数据管线驱动的人形自主系统。该语料库随算力而非人力扩展。HALO 习得坐椅子、导航到指定物体、侧身穿过半开的门、蹲伏爬过桌底、推箱子等物理技能，并协调这些技能解决复合长时程任务。训练中从未见过真实家庭，HALO 零样本迁移到物理世界，在六个未建图真实环境中 360 次试验成功率达 91%。代码和视频：https://apple-aiml-research.github.io/ml-halo/。

原文摘要

Humanoid robots have the potential to automate a wide range of physical tasks. Progress toward generalist humanoid autonomy, however, remains constrained by a data bottleneck: mapping pixels and free-form language to low-level joint commands requires paired supervision, and no corpus of such data exists at sufficiently large scale. Some approaches rely on human teleoperation and wearable capture devices, which require at least an hour of human time to produce an hour of data. We present HALO, a humanoid autonomy system powered by a pipeline that synthesizes vision-language-motion data across a thousand procedurally generated homes. The corpus scales with compute rather than human labor. HALO acquires a repertoire of physical skills, including sitting on chairs, navigating to named objects,...

#论文 #arXiv #CV #小凯