eval_taken_5

该数据集使用LeRobot创建，专为机器人学任务设计。数据集包含一个总集，总计3205帧，帧率为30 fps。机器人类型为mobileai_robot，数据以parquet文件格式存储，总数据文件大小为100 MB，视频文件大小为200 MB。特征包括动作数据（16维浮点数组，表示左右机械臂关节位置和速度）、观测状态（类似的动作数据）、三个摄像头图像（cam_high、cam_left_wrist、cam_right_wrist，均为480x640分辨率、3通道视频，使用av1编解码器），以及时间戳、帧索引、集索引、索引和任务索引等元数据。数据集仅包含训练分割（0:1），适用于机器人控制和视觉任务。This dataset was developed using LeRobot and is specifically tailored for robotics tasks. It consists of a single total dataset with 3205 total frames at a frame rate of 30 fps. The robotic platform is mobileai_robot. The data is stored in Parquet file format, with a total data file size of 100 MB and a total video file size of 200 MB. The dataset features include action data (a 16-dimensional floating-point array representing the joint positions and velocities of the left and right robotic arms), observation states (of the same format as the action data), three camera image streams (cam_high, cam_left_wrist, and cam_right_wrist, all with 480x640 resolution, 3-channel video using the AV1 codec), as well as metadata such as timestamps, frame indices, split indices, sample indices, and task indices. The dataset only contains the training split (0:1), and is suitable for robotics control and computer vision tasks.