oxe-aug/language_table_train_420000_425000_augmented

language_table_train_420000_425000_augmented数据集是一个机器人增强数据集，基于OXE数据集扩展。它包含了8种不同类型的机器人（google_robot、images、jaco、kinova3、kuka_iiwa、panda、sawyer、ur5e），总共有5000个剧集和79111帧，采用10帧/秒的帧率。数据集提供了丰富的特征字段，包括不同机器人的图像、末端执行器位置、关节位置等，并提供了自然语言指令描述任务。该数据集适用于机器人跨体型策略学习的研究。

The language_table_train_420000_425000_augmented dataset is an augmented robot dataset based on the OXE dataset. It includes 8 types of robots (google_robot, images, jaco, kinova3, kuka_iiwa, panda, sawyer, ur5e), with a total of 5,000 episodes and 79,111 frames, recorded at 10 frames per second. The dataset provides a rich set of features, including images for different robots, end-effector positions, joint positions, etc., and offers natural language instructions describing the tasks. This dataset is suitable for research on cross-embodiment policy learning for robots.