arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2410.11792cs.ROcs.AIcs.CVcs.LG

OKAMI:通过单视频模仿教人形机器人操作技能

OKAMI: Teaching Humanoid Robots Manipulation Skills through Single Video Imitation

UT Austin(德克萨斯大学奥斯汀分校);NVIDIA Research(英伟达研究院)

Jinhan Li, Yifeng Zhu, Yuqi Xie, Zhenyu Jiang, Mingyo Seo, Georgios Pavlakos, Yuke Zhu

AI总结:

本文提出OKAMI方法,通过单个RGB-D视频演示利用对象感知重定向技术教人形机器人学习操作技能,在开放世界模仿中超越基线,并利用其轨迹训练闭环策略实现79.2%的平均成功率。

摘要:

我们研究通过单视频演示模仿来教人形机器人操作技能的问题。我们引入了OKAMI,这是一种从单个RGB-D视频生成操作计划并推导出可执行策略的方法。我们方法的核心是对象感知重定向,它使人形机器人能够模仿RGB-D视频中的人类动作,同时在部署期间适应不同的物体位置。OKAMI使用开放世界视觉模型来识别任务相关对象,并分别重定向身体动作和手部姿态。我们的实验表明,OKAMI在各种视觉和空间条件下实现了强大的泛化能力,在基于观察的开放世界模仿上优于最先进的基线方法。此外,OKAMI的 rollout 轨迹被用于训练闭环视觉运动策略,这些策略在无需劳动密集型遥操作的情况下实现了79.2%的平均成功率。更多视频可在我们的网站上找到 https://ut-austin-rpl.github.io/OKAMI/。

英文摘要:

We study the problem of teaching humanoid robots manipulation skills by imitating from single video demonstrations. We introduce OKAMI, a method that generates a manipulation plan from a single RGB-D video and derives a policy for execution. At the heart of our approach is object-aware retargeting, which enables the humanoid robot to mimic the human motions in an RGB-D video while adjusting to different object locations during deployment. OKAMI uses open-world vision models to identify task-relevant objects and retarget the body motions and hand poses separately. Our experiments show that OKAMI achieves strong generalizations across varying visual and spatial conditions, outperforming the state-of-the-art baseline on open-world imitation from observation. Furthermore, OKAMI rollout trajectories are leveraged to train closed-loop visuomotor policies, which achieve an average success rate of 79.2% without the need for labor-intensive teleoperation. More videos can be found on our website https://ut-austin-rpl.github.io/OKAMI/.

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。

↑