red_marker

该数据集是一个机器人数据集，使用LeRobot工具创建，包含52个episodes和12752帧，数据以30fps的帧率采集。特征包括动作（6个浮点值，对应机器人关节位置）、观测状态（6个浮点值，同样对应关节位置）、观测图像（包括头顶和手腕视角的视频，分辨率为480x640，3通道，使用AV1编解码器），以及时间戳、帧索引、episode索引、索引和任务索引等元数据。数据集使用so_follower机器人类型，总数据文件大小为100MB，视频文件大小为200MB，数据以parquet格式存储，视频以mp4格式存储，适用于机器人任务，如运动控制和视觉感知。This dataset is a robotics dataset created using the LeRobot tool, containing 52 episodes and 12752 frames collected at a frame rate of 30fps. Features include actions (6 floating-point values corresponding to robot joint positions), observation states (6 floating-point values also for joint positions), observation images (including overhead and wrist-view videos with a resolution of 480x640, 3 channels, using the AV1 codec), as well as metadata such as timestamps, frame indices, episode indices, indices, and task indices. The dataset uses the so_follower robot type, with a total data file size of 100MB and video file size of 200MB, stored in parquet format for data and mp4 format for videos. It is suitable for robotics tasks such as motion control and visual perception.