让机器人直接从深度图像学踢球，能避人还能控球。
Vision-Based Dribbling for Humanoid Soccer via Privileged Representation Learning
- 用深度图编码器嵌入强化学习策略，实现视觉与动作一体化
- 对静态障碍物成功率96%，对抗移动对手成功率达46%
- 无需额外状态估计，适合真实场景的实时踢球任务
近期人形机器人研究强调了可部署的运动操作技能的重要性。在躲避主动对手的同时带球踢球，需要同时保持平衡、精确控球，并感知动态对手，且受限于机载传感器和实时性要求。现有方法通常将感知与运动分离，在遮挡、快速球速和复杂对手互动下易失效，因感知未针对控制优化。本文提出一种集成框架，通过特定任务投影层将时序深度编码器嵌入强化学习策略。应用于模拟的Booster T1人形机器人，证明可直接从深度观测中学习视觉驱动、对手感知的带球策略，无需显式状态估计或特权场景信息。所学策略在目标导向带球任务中达成100%成功率，单个静态障碍物下为96%，对抗主动进攻的移动对手时达46%成功率。结果表明该框架支持在正常及中等动态环境下鲁棒视觉带球，为应对更复杂移动对手场景奠定基础。
原文摘要 · Abstract (English)
Recent advances in humanoid robotics have highlighted the importance of deployable loco-manipulation skills. Dribbling a soccer ball while evading active opponents requires simultaneous balance, precise ball control, and awareness of a dynamic adversary under onboard sensing and real-time constraints. Existing approaches typically separate perception and motion, which can be effective in controlled settings but may fail under occlusions, fast ball movements, and complex opponent interactions, since perception is not directly optimized for control. We propose an integrated approach in which a temporal depth encoder is embedded into a reinforcement learning policy through a task-specific projection layer. We apply this framework to a simulated Booster T1 humanoid robot and show that it is possible to learn vision-based, opponent-aware dribbling directly from depth observations, without explicit state estimation or privileged scene information. The learned policy achieves 100% success in nominal target-driven dribbling and 96% success with a single static obstacle, while reaching 46% success against an actively moving ball-attacker opponent. These results demonstrate that the proposed framework supports robust vision-based dribbling in nominal and moderately dynamic settings, and provides a strong foundation for handling more challenging moving-adversary scenarios.
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。