individual_grasp_collect_uncoated

该数据集是一个机器人抓取数据集，使用LeRobot工具创建。它包含259个训练集，总计22091帧数据，涉及抓取拾取器机器人类型。数据集结构包括观察图像（来自腕部摄像头，分辨率为480x640，3通道，20fps）、观察状态（14维浮点数组，表示末端执行器位置、姿态等）、观察标线（2维浮点数组）、动作（7维浮点数组，表示位移和抓取器控制）、结果（1维浮点数组，表示结果代码）以及其他元数据如时间戳、帧索引、集索引等。数据以Parquet格式存储，视频以MP4格式存储，适用于机器人学习和控制任务。This dataset is a robotic grasping dataset created using the LeRobot tool. It contains 259 training sets, totaling 22,091 frames of data, involving a gripper robot type. The dataset structure includes observation images (from a wrist camera with a resolution of 480x640, 3 channels, 20fps), observation states (14-dimensional float arrays representing end-effector position, orientation, etc.), observation markers (2-dimensional float arrays), actions (7-dimensional float arrays representing displacement and gripper control), results (1-dimensional float arrays representing result codes), and other metadata such as timestamps, frame indices, and set indices. The data is stored in Parquet format, and videos are stored in MP4 format, suitable for robotic learning and control tasks.