DemoGen：用于数据高效视觉运动策略学习的合成演示生成
DemoGen: Synthetic Demonstration Generation for Data-Efficient Visuomotor Policy Learning
AI总结：
DemoGen提出一种低成本的合成演示生成方法，仅需单个人工演示即可通过空间增强和3D编辑生成多样化演示，显著提升视觉运动策略在多种真实操作任务中的性能，并支持干扰抵抗和障碍物避免等额外能力。
摘要：
视觉运动策略在机器人操作中展现出巨大潜力，但通常需要大量人工收集的数据才能实现有效性能。数据需求的一个关键原因是其有限的空间泛化能力，这要求在不同物体配置下进行广泛的数据收集。在这项工作中，我们提出了DemoGen，一种低成本、全合成的自动演示生成方法。仅使用每个任务一个人工收集的演示，DemoGen通过将演示的动作轨迹适应到新的物体配置来生成空间增强的演示。视觉观察通过利用3D点云作为模态并通过3D编辑重新排列场景中的主体来合成。实验上，DemoGen在多种真实世界操作任务中显著提升了策略性能，显示出其在涉及可变形物体、灵巧手末端执行器和双臂平台等挑战性场景中的适用性。此外，DemoGen可以扩展以实现额外的分布外能力，包括干扰抵抗和障碍物避免。
英文摘要：
Visuomotor policies have shown great promise in robotic manipulation but often require substantial amounts of human-collected data for effective performance. A key reason underlying the data demands is their limited spatial generalization capability, which necessitates extensive data collection across different object configurations. In this work, we present DemoGen, a low-cost, fully synthetic approach for automatic demonstration generation. Using only one human-collected demonstration per task, DemoGen generates spatially augmented demonstrations by adapting the demonstrated action trajectory to novel object configurations. Visual observations are synthesized by leveraging 3D point clouds as the modality and rearranging the subjects in the scene via 3D editing. Empirically, DemoGen significantly enhances policy performance across a diverse range of real-world manipulation tasks, showing its applicability even in challenging scenarios involving deformable objects, dexterous hand end-effectors, and bimanual platforms. Furthermore, DemoGen can be extended to enable additional out-of-distribution capabilities, including disturbance resistance and obstacle avoidance.
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。
