DRMNmadhan/record_2

该数据集是一个用于机器人学任务的数据集，专门针对so101_follower类型机器人，由LeRobot工具创建。数据集包含4个训练片段，总计2392帧，帧率为30fps。数据以Parquet格式存储，并包含视频文件（MP4格式）。特征包括：动作数据（6维浮点数组，表示机械臂的肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置）、观测状态（与动作相同的6维关节位置）、夹爪摄像头图像（480x640分辨率，3通道）和全局摄像头图像（720x1280分辨率，3通道），以及时间戳、帧索引、片段索引等元数据。数据集主要用于机器人控制与视觉感知的机器学习研究。This dataset is designed for robotics tasks, specifically for the so101_follower robot type, created using the LeRobot tool. It contains 4 training episodes with a total of 2392 frames at 30 fps. The data is stored in Parquet format and includes video files (MP4). Features consist of: action data (6-dimensional float32 array representing robot arm joint positions: shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper position), observation state (same 6-dimensional joint positions), gripper camera images (480x640 resolution, 3 channels), global camera images (720x1280 resolution, 3 channels), along with metadata such as timestamp, frame index, and episode index. The dataset is intended for machine learning research in robot control and visual perception.