Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition
Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition
复制标题
DOI：
10.48550/arxiv.2307.14535
复制
发表时间：
2023-07
期刊：
影响因子：
--
通讯作者：
Huy Ha;Peter R. Florence;Shuran Song
中科院分区：
文献类型：
--
作者：
Huy Ha;Peter R. Florence;Shuran Song
我们提出了一个机器人技能获取的框架，它1)有效地放大语言标记的机器人数据的数据生成，2)有效地将这些数据提取到一个健壮的多任务语言条件视觉-运动策略中。对于(1)，我们使用大型语言模型(LLM)来指导高层规划，并使用基于采样的机器人规划器(例如运动采样器或抓取采样器)来生成多样化和丰富的操作轨迹。为了使这一数据收集过程更具实用性，LLM还为每项任务的成功条件推断代码片段，同时使数据收集过程能够检测失败和重试，并自动标记成功/失败的轨迹。对于(2)，我们将扩散策略单任务行为克隆方法扩展到具有语言条件作用的多任务设置。最后，我们提出了一个新的多任务基准，包括五个领域的18个任务，以测试长期行为、常识性推理、工具使用和直观物理。我们发现，我们的提炼策略在其数据收集过程中成功地学习了健壮的重试行为，同时在五个领域中将绝对成功率平均提高了33.2%。Https://www.cs.columbia.edu/~huy/scalingup/.上提供了代码、数据和其他定性结果
We present a framework for robot skill acquisition, which 1) efficiently scale up data generation of language-labelled robot data and 2) effectively distills this data down into a robust multi-task language-conditioned visuo-motor policy. For (1), we use a large language model (LLM) to guide high-level planning, and sampling-based robot planners (e.g. motion or grasp samplers) for generating diverse and rich manipulation trajectories. To robustify this data-collection process, the LLM also infers a code-snippet for the success condition of each task, simultaneously enabling the data-collection process to detect failure and retry as well as the automatic labeling of trajectories with success/failure. For (2), we extend the diffusion policy single-task behavior-cloning approach to multi-task settings with language conditioning. Finally, we propose a new multi-task benchmark with 18 tasks across five domains to test long-horizon behavior, common-sense reasoning, tool-use, and intuitive physics. We find that our distilled policy successfully learned the robust retrying behavior in its data collection procedure, while improving absolute success rates by 33.2% on average across five domains. Code, data, and additional qualitative results are available on https://www.cs.columbia.edu/~huy/scalingup/.