一个策略让机器人学会多种动态动作，还能长时间稳定运行。
KungfuBot2: Learning Versatile Motion Skills for Humanoid Whole-Body Control
- 用混合专家架构让不同动作各司其职，同时保持泛化能力。
- 通过分段追踪奖励，实现分钟级长序列稳定运动。
- 适合想构建多功能人形机器人的研究者和开发者。
通过跟踪多种人类动作来学习通用的全身运动技能，是实现通用人形机器人的关键一步。这一任务极具挑战性，因为单一策略需掌握广泛的动作技能，并在长时序序列中保持稳定性。为此，我们提出VMS，一种统一的全身控制器，使机器人能在单个策略内学习多样且动态的行为。框架融合了混合追踪目标，平衡局部动作保真度与全局轨迹一致性，并采用正交专家混合（OMoE）结构，促进技能专精同时增强跨动作泛化能力。进一步引入分段级追踪奖励，放宽逐步匹配要求，提升对全局位移和瞬时误差的鲁棒性。我们在仿真与真实世界实验中全面验证VMS，展示了对动态技能的精准模仿、分钟级序列的稳定表现，以及对未见动作的强大泛化能力。这些结果表明VMS可作为可扩展的人形全身控制基础。项目页面见https://kungfubot2-humanoid.github.io。
原文摘要 · Abstract (English)
Learning versatile whole-body skills by tracking various human motions is a fundamental step toward general-purpose humanoid robots. This task is particularly challenging because a single policy must master a broad repertoire of motion skills while ensuring stability over long-horizon sequences. To this end, we present VMS, a unified whole-body controller that enables humanoid robots to learn diverse and dynamic behaviors within a single policy. Our framework integrates a hybrid tracking objective that balances local motion fidelity with global trajectory consistency, and an Orthogonal Mixture-of-Experts (OMoE) architecture that encourages skill specialization while enhancing generalization across motions. A segment-level tracking reward is further introduced to relax rigid step-wise matching, enhancing robustness when handling global displacements and transient inaccuracies. We validate VMS extensively in both simulation and real-world experiments, demonstrating accurate imitation of dynamic skills, stable performance over minute-long sequences, and strong generalization to unseen motions. These results highlight the potential of VMS as a scalable foundation for versatile humanoid whole-body control. The project page is available at https://kungfubot2-humanoid.github.io.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。