b601_towel_fold
收藏资源简介：
该数据集是使用LeRobot框架收集的机器人毛巾折叠任务演示数据。包含18个演示片段（episodes），共计57829帧，所有帧对应单一任务（毛巾折叠）。数据特征包括：机器人7个关节的连续动作（位置）、21维状态观测（位置、力矩、速度），以及来自左眼、右眼和腕部摄像头的RGB图像（帧率30fps，分辨率分别为672x376和640x360）。此外，还包含时间戳、帧索引、episode索引等元数据。适用于模仿学习、机器人操作等研究。
This dataset contains robot towel folding task demonstration data collected using the LeRobot framework. It includes 18 demonstration episodes, totaling 57,829 frames, all corresponding to a single task (towel folding). Data features include continuous actions (positions) of 7 robot joints, 21-dimensional state observations (positions, torques, velocities), and RGB images from left-eye, right-eye, and wrist cameras (frame rate 30fps, resolutions 672x376 and 640x360 respectively). Additionally, metadata such as timestamps, frame indices, and episode indices are included. Suitable for imitation learning, robotic manipulation, and other research.
