Simulation and Real Robot Trajectories
收藏arXiv2025-09-30 收录
下载链接:
https://ut-austin-rpl.github.io/olaf/
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含了针对各种操作任务的演示和机器人轨迹数据,这些任务既包括了模拟环境下的操作,也包括了真实机器人互动的情况。在此过程中,用户提供了口头更正。此外,数据集还包含了通过键盘输入的口头反馈,并为未来可能的语音识别系统应用预留了空间。在规模上,数据集包含了M+N条轨迹,其中模拟环境下有M=50条和N=100条,真实环境下有M=40条和N=80条。这些数据集针对的是带有口头更正的长视距操作任务。
This dataset contains demonstrations and robotic trajectory data for diverse manipulation tasks, encompassing both simulated environment-based operations and real robotic interaction scenarios. During the data acquisition process, users provided verbal corrections. Additionally, the dataset includes verbal feedback entered via keyboard, and reserves space for potential future applications of speech recognition systems. In terms of scale, the dataset consists of M+N total trajectories: there are M=50 and N=100 trajectories for simulated environments, and M=40 and N=80 trajectories for real-world scenarios. These datasets are tailored for long-horizon manipulation tasks with verbal corrections.
提供机构:
University of Texas at Austin



