oxe-aug/berkeley_autolab_ur5_train_300_400_augmented

berkeley_autolab_ur5_train_300_400是一个机器人学数据集，包含多种机器人的增强图像、末端执行器位置、关节角度等信息。数据集由100个剧集组成，共有9934帧，分为900个视频。数据集适用于机器人政策的跨体型扩展学习。berkeley_autolab_ur5_train_300_400 is a robotics dataset that includes augmented images for various robots, end-effector positions, joint angles, and more. The dataset consists of 100 episodes with a total of 9,934 frames, split into 900 videos. It is suitable for cross-embodiment policy learning in robotics.