chomeed/ai_worker_board_handover_dagger_arm_only_failure

该数据集是一个机器人相关数据集，使用LeRobot创建，专门用于机器人学习和控制任务。数据集包含9个episodes，总计2616帧，帧率为30fps，所有数据均用于训练。数据以parquet格式存储，总数据文件大小为100MB，视频文件大小为200MB。数据集的结构包括多个关键特征：观测状态（observation.state）为38维浮点数组，表示左臂和右臂各7个关节以及左右夹爪的关节状态；动作（action）为16维浮点数组，控制左臂和右臂各7个关节以及左右夹爪；图像观测包括头戴摄像头（分辨率720x1280）、左腕摄像头（分辨率240x424）和右腕摄像头（分辨率240x424）的视频数据，均为RGB格式（3通道）。此外，还包含时间戳、帧索引、episode索引、索引和任务索引等元数据。该数据集适用于机器人模仿学习、强化学习等研究，重点关注双臂机器人的手部交接任务（handover），但仅包含失败案例（arm_only_failure）。数据集遵循Apache 2.0许可证。This dataset is a robotics-related dataset created using LeRobot, specifically designed for robot learning and control tasks. It contains 9 episodes with a total of 2616 frames at 30fps, all used for training. The data is stored in parquet format, with a total data file size of 100MB and video file size of 200MB. The dataset structure includes key features: observation.state is a 38-dimensional float32 array representing joint states for the left arm (7 joints), right arm (7 joints), and left/right grippers; action is a 16-dimensional float32 array controlling the left arm (7 joints), right arm (7 joints), and left/right grippers; image observations include video data from a head-mounted camera (resolution 720x1280), left wrist camera (resolution 240x424), and right wrist camera (resolution 240x424), all in RGB format (3 channels). Additionally, it contains metadata such as timestamp, frame index, episode index, index, and task index. This dataset is suitable for research in robot imitation learning, reinforcement learning, etc., focusing on bimanual robot handover tasks but only includes failure cases (arm_only_failure). The dataset is licensed under Apache 2.0.