外观
论文arXiv 2609.341991 个月前
WB-WAM：用于类人机器人移动操作的异构身体-手部预训练
原文链接：WB-WAM: Heterogeneous Body-Hand Pre-training for Humanoid Loco-Manipulation（arXiv 2609.34199 · 发布日期：2026-09-28）
WB-WAM: Heterogeneous Body-Hand Pre-training for Humanoid Loco-Manipulation
WB-WAM通过全身动作监督的生成式视频预训练，解决类人移动操作协同问题，并在仿真和真实任务上验证。
核心要点
- 动机是类人移动操作需要身体与手协同，而常规预训练数据对全身运动覆盖有限。
- 方法将身体、根部和灵巧手标注纳入共享物理动作空间，联合学习视频与动作。
- 模型经PICO中期训练和前向运动学辅助监督适配机器人任务，并利用人类示教补充机器人数据。
- 仿真HumanoidArena成功率为81.9%，真实世界五项任务平均成功率为84.0%。
验证范围与局限：验证范围包括仿真HumanoidArena和真实世界五项任务。
英文摘要 (Abstract)
Humanoid loco-manipulation demands coordinated body and hand behavior, while conventional robot pre-training data provide limited coverage of such whole-body motion. We present WB-WAM, a World Action Model that incorporates explicit whole-body action supervision into generative video pre-training. A shared physical action space integrates body, root, and dexterous hand annotations from heterogeneous sources, enabling joint video and action learning from 1880.2 hours of partially annotated video and motion data. The resulting priors are refined through PICO mid-training and adapted to robot tasks with auxiliary forward kinematics supervision. We construct WB-Datasets to support these stages with retargeted egocentric human demonstrations and robot trajectories, allowing task-aligned human motion to supplement limited robot data. Evaluations in simulation demonstrate strong whole-body task performance with 81.9% in HumanoidArena, while real-world experiments further validate WB-WAM with 84.0% mean success across five tasks. Moreover, task-aligned PICO mid-training improves downstream task performance while reducing the need for real-robot demonstrations. These results support heterogeneous whole-body pre-training and human motion transfer as a practical route to data-efficient humanoid loco-manipulation.
作者：Chuan Qin, Shaoting Zhu, Siyuan Luo, Siqiao Huang, Hongyu Zhao, Shanaka Kristombu Baduge, Hang Zhao
来源
本页事实对应的公开页面，以原文为准。