CMU Stretch
收藏资源简介:
CMU Stretch 数据集由卡内基梅隆大学(CMU)的研究团队创建,专注于机器人在真实环境中的操作任务,如开门、拉抽屉等。该数据集包含在模拟环境和真实世界中采集的数据,旨在促进机器人对可折叠物体的操作与控制研究。数据集包含 54K 图像-轨迹-接触点元组,每个元组记录了机器人在执行任务时的图像、轨迹和接触点信息。数据来源于 EpicKitchens-100 视频数据集,涉及多种厨房任务和操作。
The CMU Stretch Dataset was developed by a research team at Carnegie Mellon University (CMU), focusing on real-world robotic manipulation tasks such as opening doors and pulling drawers. The dataset includes data collected from both simulated environments and real-world scenarios, aiming to advance research on robotic manipulation and control of foldable objects. It contains 54K image-trajectory-contact point tuples, with each tuple recording the robot's captured images, movement trajectory, and contact point information during task execution. The data is sourced from the EpicKitchens-100 video dataset, covering a wide range of kitchen-related tasks and manipulation operations.




