oxe-auge/fractal20220817_data_train_62500_65000_augmented

该数据集是一个用于机器人学习和增强学习任务的大规模数据集，包含多种机器人的图像和状态信息。数据集提供了2500个episodes和108654帧的数据，帧率为3fps。数据集的特征包括机器人的图像（256×320×3）、关节位置、末端执行器位置、自然语言指令和任务索引等。数据集支持多种机器人，如jaco、kinova3、kuka_iiwa、panda、sawyer、ur5e、xarm7等。数据集的使用需要引用原始数据集和OXE-AugE数据集的论文。

This dataset is a large-scale dataset for robotics learning and reinforcement learning tasks, containing images and state information of multiple robots. The dataset provides 2,500 episodes and 108,654 frames of data with a frame rate of 3fps. The features of the dataset include robot images (256×320×3), joint positions, end-effector positions, natural language instructions, and task indices. The dataset supports multiple robots such as jaco, kinova3, kuka_iiwa, panda, sawyer, ur5e, xarm7, etc. The use of the dataset requires citation of both the original dataset and the OXE-AugE dataset paper.