edmos7/so101_test-trajectory

该数据集是一个用于机器人学习的轨迹数据集，由LeRobot创建。它包含15个episodes和10910帧数据，涉及单个任务。数据以parquet格式存储，包括机器人动作（6维浮点数组，对应关节控制）、状态观测（6维浮点数组，表示关节位置）、时间戳、帧索引、episode索引等。此外，数据集提供视觉数据：来自MAIN_CAM和SEC_CAM的视频流（分辨率480x640，30fps，AV1编码），以及轨迹图像。这些数据用于机器人控制任务，支持机器人学研究和应用。

This dataset is a trajectory dataset for robot learning, created by LeRobot. It contains 15 episodes and 10910 frames, involving a single task. The data is stored in parquet format, including robot actions (6-dimensional float32 arrays for joint control), state observations (6-dimensional float32 arrays for joint positions), timestamps, frame indices, episode indices, etc. Additionally, the dataset provides visual data: video streams from MAIN_CAM and SEC_CAM (resolution 480x640, 30fps, AV1 codec), and trajectory images. This data is used for robot control tasks, supporting robotics research and applications.