so101_desk_trash3

该数据集是通过LeRobot创建的机器人学习数据集，专注于机器人控制任务。它包含多个特征，包括动作、观察状态（如关节位置）、前摄像头图像（分辨率为480x640的RGB视频）、时间戳、帧索引、剧集索引、索引和任务索引。具体特征涉及6个自由度的关节位置（例如肩部平移、肩部升降、肘部弯曲、手腕弯曲、手腕旋转和夹爪位置）。数据以Parquet文件格式存储，视频以MP4格式存储，帧率为30fps。数据集旨在支持机器人模仿学习或强化学习任务，用于分析和建模机器人动作与状态。

This dataset is created by LeRobot for robot learning, focusing on robot control tasks. It includes multiple features such as action, observation state (e.g., joint positions), front camera images (RGB video with a resolution of 480x640), timestamp, frame index, episode index, index, and task index. Specific features involve 6 degrees of freedom joint positions (e.g., shoulder translation, shoulder elevation, elbow flexion, wrist flexion, wrist rotation, and gripper position). Data is stored in Parquet file format, with videos in MP4 format at a frame rate of 30fps. The dataset is designed to support robot imitation learning or reinforcement learning tasks, aiming to analyze and model robot actions and states.