Smooth Action Chunking Via Movement-Primitive Representations for Imitation Learning
作者: Shota Nomura, Takayuki Osa, Yasuhiro Kato, Kazuki Ota, Naoki Shitanda, Tatsuya Harada
分类: cs.LG, cs.RO
发布日期: 2026-10-29
💡 一句话要点
通过运动原语表示提出平滑动作分块以解决模仿学习中的抖动问题
🎯 匹配领域: 支柱一：机器人控制 (Robot Control) 支柱二：RL算法与架构 (RL & Architecture) 支柱九：具身大模型 (Embodied Foundation Models)
关键词: 模仿学习 动作分块 运动原语 轨迹平滑 机器人控制 自动驾驶 人机交互
📋 核心要点
- 现有模仿学习方法在动作分块时，轨迹的时间结构仅隐式学习，导致指令轨迹抖动明显。
- 本文提出通过运动原语表示来构建平滑轨迹，确保动作分块的平滑性，同时不改变现有策略架构。
- 在实验中，该方法在一台双手ALOHA平台上显著降低了块内抖动，并有效减少了边界不连续性。
📝 摘要（中文）
模仿学习因其概念简单、可扩展性强以及能够利用大规模数据集而广泛应用于基于视觉的机器人策略。随着这些策略能力的提升，动作分块成为其标准接口。然而，现有方法在时间结构上仅隐式学习，导致指令轨迹在块内和块间都出现抖动。本文通过将运动原语基础函数引入动作分块，构建平滑轨迹，同时保持策略架构和训练目标不变。该方法在一台双手ALOHA平台上进行评估，显著降低了块内抖动，并有效减少了边界不连续性。
🔬 方法详解
问题定义：本文旨在解决模仿学习中动作分块导致的轨迹抖动问题。现有方法在时间结构上仅隐式学习，导致指令轨迹在块内和块间都出现明显的抖动，影响机器人运动的平滑性。
核心思路：论文提出将运动原语基础函数引入动作分块，通过线性组合构建平滑轨迹。这种设计确保了轨迹的平滑性，同时不改变现有的策略架构、训练目标和推理过程。
技术框架：整体架构包括动作分块模块和运动原语表示模块。动作分块模块负责生成未来动作，而运动原语表示模块则通过平滑基础函数生成平滑轨迹。推理时，预测的系数被条件化，以确保新块通过上一个块的最后指令点。
关键创新：最重要的创新在于引入运动原语表示，使得轨迹生成过程平滑化，显著降低了块内抖动，并且提供了一种闭式推理时间修正方法，避免了重新训练的需求。
关键设计：在参数设置上，使用平滑基础函数的线性组合来表示轨迹段。损失函数保持不变，网络结构也未做大幅调整，确保了与现有方法的兼容性。
🖼️ 关键图片
📊 实验亮点
实验结果表明，所提方法在一台双手ALOHA平台上，块内抖动降低了一个数量级，而在迭代基础上降低了2.4倍。边界不连续性减少了73-84%，且每步的计算成本几乎可以忽略不计。
🎯 应用场景
该研究的潜在应用领域包括机器人控制、自动驾驶和人机交互等。通过提升机器人运动的平滑性，可以显著改善其在复杂环境中的表现，未来可能推动更高效的自动化系统和智能机器人技术的发展。
📄 摘要（原文）
Imitation learning has become a widely adopted approach for vision-based robot policies, owing to its conceptual simplicity, scalability, and ability to leverage large-scale datasets. As these policies have grown more capable, action chunking has become their standard interface. At each decision step the policy outputs several future actions at once. Backbones from diffusion policies to large vision-language-action models all share it. Whatever backbone produces the chunk, its temporal structure is learned only implicitly. The commanded trajectory is therefore jerky both within a chunk and between successive chunks. On real hardware this appears as visible vibration and as repeated saturation of the follower's velocity limit. We address this issue by incorporating a movement-primitive-based trajectory representation into action chunking. Specifically, a trajectory segment is represented as a linear combination of smooth basis functions, yielding smooth trajectories by construction while leaving the policy architecture, training objective, and inference procedure unchanged. The same representation also yields a closed-form, inference-time correction for the second source of jerk. We condition the predicted coefficients so that the new chunk passes through a via-point at the previous chunk's last commanded action. This correction needs no retraining and, unlike asynchronous Real-Time Chunking, applies to policies queried in one shot. We instantiate it on a one-shot (ACT) and an iterative (DiT) backbone and evaluate in simulation and on a real bimanual ALOHA platform. The representation lowers within-chunk jerk by an order of magnitude on the one-shot backbone and by 2.4 times on the iterative one, with task success statistically non-inferior. The conditioning then removes 73-84% of the boundary discontinuity, at negligible per-step cost.