CarolinePascal/so101_audio

该数据集由LeRobot创建，专注于机器人技术领域。数据集包含1个总任务和600帧数据，帧率为30fps。数据结构丰富，包括6维浮点动作数据（如肩部、肘部、手腕等位置）、6维浮点观察状态数据、前部和腕部摄像头视频数据（480x640分辨率，3通道，30fps）以及单通道音频数据（采样率48kHz）。数据以parquet格式存储，视频和音频分别以mp4和m4a格式存储。

This dataset was created by LeRobot and focuses on the field of robotics. It contains 1 total task and 600 frames of data with a frame rate of 30fps. The dataset structure is rich, including 6-dimensional floating-point action data (such as shoulder, elbow, wrist positions), 6-dimensional floating-point observation state data, front and wrist camera video data (480x640 resolution, 3 channels, 30fps), and single-channel audio data (sampling rate 48kHz). The data is stored in parquet format, with videos and audio stored in mp4 and m4a formats respectively.