oxe-auge/fractal20220817_data_train_50000_52500_augmented

fractal20220817_data_train_50000_52500_augmented是一个用于机器人学习和增强现实控制的大规模数据集。数据集包含2500个episodes和107,972帧数据，涵盖了多种机器人（如jaco、kinova3、kuka_iiwa、panda、sawyer、ur5e、xarm7）的图像和状态信息。数据集的特征包括256×320×3分辨率的机器人图像、关节位置、末端执行器位置、自然语言指令等。数据以parquet格式存储，并提供了详细的数据布局和特征描述。

fractal20220817_data_train_50000_52500_augmented is a large-scale dataset for robotics learning and augmented reality control. The dataset contains 2,500 episodes and 107,972 frames, covering images and state information of various robots (e.g., jaco, kinova3, kuka_iiwa, panda, sawyer, ur5e, xarm7). Features of the dataset include 256×320×3 resolution robot images, joint positions, end-effector positions, natural language instructions, etc. The data is stored in parquet format and comes with detailed data layout and feature descriptions.