dice4-trajectory

该数据集是使用LeRobot创建的机器人控制数据集，专门用于机器人学习任务。数据集包含108个完整的情节（episodes），总计55590帧，涵盖2个不同的任务。数据以parquet格式存储，总数据文件大小为100MB，视频文件大小为200MB。帧率为30fps。数据集的特征包括：6维动作空间（控制机器人的肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪动作），6维状态观测（与动作空间相同的关节状态），时间戳、帧索引、情节索引和任务索引。观测部分包括来自MAIN_CAM和SEC_CAM的视频数据（分辨率480x640，3通道），以及轨迹图像观测。机器人类型为so100，数据按块组织，每个块大小为1000帧。数据集仅包含训练分割（所有108个情节）。

This dataset is a robot control dataset created using LeRobot, specifically designed for robot learning tasks. It contains 108 complete episodes, totaling 55,590 frames and covering 2 different tasks. The data is stored in parquet format, with a total data file size of 100MB and a video file size of 200MB. The frame rate is 30fps. Features of the dataset include: a 6-dimensional action space (controlling the robots shoulder translation, shoulder lift, elbow bend, wrist bend, wrist rotation, and gripper action), a 6-dimensional state observation (joint states identical to the action space), timestamps, frame indices, episode indices, and task indices. Observations include video data from MAIN_CAM and SEC_CAM (resolution 480x640, 3 channels), as well as trajectory image observations. The robot type is so100, and the data is organized in chunks, with each chunk containing 1000 frames. The dataset only includes a training split (all 108 episodes).