发表机构
Zhejiang University; Hong Kong Embodied AI Lab; Mondo Robotics; The Chinese University of Hong Kong（浙江大学; 香港具身智能实验室; Mondo Robotics; 香港中文大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对人形全身示范收集难的问题，提出WB-UMI，通过共享末端执行器接口解耦全身协调与任务语义学习，结合扩散策略与异步层级控制，在G1上实现四任务实时闭环迁移，成功率最高达90%。
AI 中文摘要
收集人形机器人全身操作的示范数据主要依赖遥操作，这种方式成本高昂且难以规模化。通用操作接口（UMI）提供了一种可扩展的数据收集范式，但仅凭末端执行器轨迹无法确定人形机器人的全身协调，这不足以支持全身示范数据的收集。为此，我们提出了全身UMI（WB-UMI），一种任务无关、实时且由末端执行器条件驱动的运动生成器，通过共享的末端执行器接口将全身协调学习与任务语义学习解耦。扩散策略从原生UMI示范中学习，而WB-UMI则独立地从重定向的运动捕捉数据中学习，在特定任务的数据收集过程中无需身体追踪器或配对的图像-全身示范。在实际部署中，一个异步层级结构集成了扩散策略、运动生成器以及带有延迟补偿和实测状态反馈的全身控制器。在G1上的真实机器人实验支持跨四个任务的实时闭环迁移，在抽屉关闭任务中达到90%的成功率，在货架取放任务中达到80%，在抛球任务中达到30%，在Loco-PnP任务中达到40%，这证明了该层级结构在将原生UMI技能迁移至人形全身操作方面的有效性。
英文摘要
Collecting whole-body demonstrations for humanoid manipulation mostly relies on teleoperation, which is costly and hard to scale up. The Universal Manipulation Interface (UMI) provides a scalable data collection paradigm, but end-effector trajectories alone underdetermine humanoid whole-body coordination, which is insufficient for whole-body demonstration collection. Therefore, we introduce Whole-Body UMI (WB-UMI), a task-agnostic, real-time and end-effector conditioned motion generator that decouples whole-body coordination learning from task semantics learning through a shared end-effector interface. A diffusion policy learns from native UMI demonstrations, while WB-UMI learns independently from retargeted motion capture, requiring no body trackers or paired image--whole-body demonstrations during task-specific data collection. In real deployment, an asynchronous hierarchy integrates the diffusion policy, motion generator, and a whole-body controller with latency compensation and measured-state feedback. Real-robot experiments on G1 support real-time closed-loop transfer across four tasks, achieving 90% success in drawer closing, 80% in shelf pick-and-place, 30% in ball toss, and 40% in Loco-PnP, which shows the effectiveness of this hierarchy in transferring native UMI skills to humanoid whole-body manipulation.
Comments8 pages, 5 figures
