从专家模型到通用模型：迈向人形机器人通用全身控制
From Experts to a Generalist: Toward General Whole-Body Control for Humanoid Robots
AI总结：
针对人形机器人全身控制泛化难的问题，提出结合运动聚类与仿真到现实迁移的BumbleBee框架，经多组专家策略训练蒸馏为通用控制器，实现了当前最优的通用全身控制效果。
摘要：
由于运动需求多样且存在数据冲突，在人形机器人上实现通用的敏捷全身控制仍是一项重大挑战。现有框架虽擅长训练针对单一运动的策略，但由于控制要求冲突和数据分布不匹配，它们难以在高度多样化的行为间实现泛化。本研究中，我们提出BumbleBee（BB，大黄蜂）——一种结合运动聚类与仿真到现实迁移的专家-通用学习框架，以攻克这些难题。BB首先利用基于自编码器的聚类方法，通过运动特征和运动描述将行为相似的运动分组。随后在每个聚类内训练专家策略，并通过迭代增量动作建模使用真实世界数据对其进行优化，以弥合仿真与现实的差距。最后，这些专家模型被蒸馏为统一的通用控制器，可在所有运动类型中保持敏捷性与鲁棒性。在两个仿真环境和一台真实人形机器人上开展的实验表明，BB实现了当前最优的通用全身控制效果，为现实世界中人形机器人的敏捷、鲁棒及可泛化性能设立了新基准。项目网页位于https://beingbeyond.github.io/BumbleBee/。
英文摘要：
Achieving general agile whole-body control on humanoid robots remains a major challenge due to diverse motion demands and data conflicts. While existing frameworks excel in training single motion-specific policies, they struggle to generalize across highly varied behaviors due to conflicting control requirements and mismatched data distributions. In this work, we propose BumbleBee (BB), an expert-generalist learning framework that combines motion clustering and sim-to-real adaptation to overcome these challenges. BB first leverages an autoencoder-based clustering method to group behaviorally similar motions using motion features and motion descriptions. Expert policies are then trained within each cluster and refined with real-world data through iterative delta action modeling to bridge the sim-to-real gap. Finally, these experts are distilled into a unified generalist controller that preserves agility and robustness across all motion types. Experiments on two simulations and a real humanoid robot demonstrate that BB achieves state-of-the-art general whole-body control, setting a new benchmark for agile, robust, and generalizable humanoid performance in the real world. The project webpage is available at https://beingbeyond.github.io/BumbleBee/.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
