Lisette1231/20260426_pickeggtoplate6

该数据集是使用LeRobot创建的机器人数据集，专为机器人学习和控制任务设计。数据集基于seeed_b601_dm_follower机器人类型，包含4个完整episodes，总计2377帧数据，覆盖1个任务。数据以parquet文件格式存储，总数据文件大小为100MB，视频文件大小为200MB，帧率为30fps。特征包括动作（7维浮点数，表示机器人关节位置如肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部偏航、腕部滚动和夹爪位置）、观测状态（同样为7维关节位置）、图像观测（来自手腕和前置摄像头的视频，分辨率480x640，3通道RGB，无音频，使用av1编解码器）以及补充信息（如策略动作、干预状态和状态指示）。数据集结构支持训练分割（episodes 0到4），适用于机器人模仿学习、强化学习或行为克隆等应用。

This dataset is a robotics dataset created using LeRobot, designed for robot learning and control tasks. It is based on the seeed_b601_dm_follower robot type, containing 4 complete episodes with a total of 2377 frames and covering 1 task. The data is stored in parquet format, with a total data file size of 100MB and video file size of 200MB, at a frame rate of 30fps. Features include actions (7-dimensional float32 representing robot joint positions such as shoulder pan, shoulder lift, elbow flex, wrist flex, wrist yaw, wrist roll, and gripper position), observation states (similarly 7-dimensional joint positions), image observations (videos from wrist and front cameras, resolution 480x640, 3-channel RGB, no audio, using av1 codec), and complementary information (e.g., policy actions, intervention status, and state indicators). The dataset structure supports a training split (episodes 0 to 4) and is suitable for applications such as imitation learning, reinforcement learning, or behavior cloning in robotics.