外观
论文arXiv 2602.066438 个月前
HuMI框架实现无机器人数据采集,人形全身操作成功率达70%
本文由 AI 根据公开内容整理,可能有误,请以原文为准:Humanoid Manipulation Interface: Humanoid Whole-Body Manipulation from Robot-Free Demonstrations(arXiv 2602.06643)
核心内容
针对当前人形机器人全身操作依赖遥操作或视觉Sim-to-RL导致硬件后勤繁重及奖励工程复杂的问题,arXiv论文提出Humanoid Manipulation Interface (HuMI)框架。该方案通过便携式硬件捕捉人体全身运动,替代传统机器人真机演示,构建了从人类动作到机器人灵巧技能的分层学习管线。
在涵盖跪姿、深蹲、抛掷、行走及双手操作的5项全身任务实验中,HuMI展现出显著的数据效能优势。量化结果表明,其数据采集效率达到传统遥操作方式的3倍,且在训练时未包含的陌生环境中,系统仍能保持70%的任务执行成功率,验证了该框架在多样化场景下的泛化能力与实用性。
关键看点
- 无机器人采集:利用便携硬件捕获全身运动数据,摆脱对真机遥操作的硬件依赖
- 数据效率倍增:分层学习管线将人体动作转化为机器人技能,采集效率达遥操作3倍
- 跨场景泛化:在跪姿、双手操作等5类全身任务测试中,未见环境成功率稳定达70%
工程与产业观察
HuMI验证了“人体动捕+分层策略”在人形机器人全身操作数据获取上的工程可行性。相比高昂的真机遥操作,这种低成本、高效率的数据范式有望缓解具身智能高质量数据稀缺瓶颈,为复杂动态环境下的灵巧操作算法训练提供标准化基础设施参考。
英文摘要 (Abstract)
Current approaches for humanoid whole-body manipulation, primarily relying on teleoperation or visual sim-to-real reinforcement learning, are hindered by hardware logistics and complex reward engineering. Consequently, demonstrated autonomous skills remain limited and are typically restricted to controlled environments. In this paper, we present the Humanoid Manipulation Interface (HuMI), a portable and efficient framework for learning diverse whole-body manipulation tasks across various environments. HuMI enables robot-free data collection by capturing rich whole-body motion using portable hardware. This data drives a hierarchical learning pipeline that translates human motions into dexterous and feasible humanoid skills. Extensive experiments across five whole-body tasks--including kneeling, squatting, tossing, walking, and bimanual manipulation--demonstrate that HuMI achieves a 3x increase in data collection efficiency compared to teleoperation and attains a 70% success rate in unseen environments.
作者:Ruiqian Nai, Boyuan Zheng, Junming Zhao, Haodong Zhu, Sicong Dai, Zunhao Chen, Yihang Hu, Yingdong Hu, Tong Zhang, Chuan Wen, Yang Gao
来源
本页事实对应的公开页面,以原文为准。