HiloHilo/makermods_pick-up-paper-cup
收藏资源简介:
该数据集使用LeRobot创建,专门用于机器人学习任务,特别是针对so101_follower机器人类型。数据集包含50个episodes,总计37454帧,帧率为30fps,所有数据用于训练。数据以分块形式存储,每个块大小为1000帧,总数据文件大小为100MB,视频文件大小为500MB。数据集的特征包括:动作数据(6个浮点数,表示肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置)、状态观测数据(与动作相同的6个关节位置)、三个摄像头的视频观测(hand_cam、front_cam和side_cam,每个视频分辨率为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。这些数据适用于机器人控制、模仿学习或强化学习任务。
This dataset was created using LeRobot and is designed for robotics learning tasks, specifically for the so101_follower robot type. It contains 50 episodes with a total of 37,454 frames at 30 fps, all allocated for training. The data is stored in chunks of 1,000 frames each, with a total data file size of 100 MB and video file size of 500 MB. Features include: action data (6 float32 values for shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper positions), observation state data (same 6 joint positions as actions), video observations from three cameras (hand_cam, front_cam, and side_cam, each with 480x640 resolution, 3 channels, using av1 codec), and metadata such as timestamp, frame index, episode index, index, and task index. This dataset is suitable for robot control, imitation learning, or reinforcement learning tasks.




