2026-07-24_shakeitup7_accelero_10kHz_nfft_512

该数据集是一个机器人学领域的数据集，使用LeRobot创建。它包含机器人动作和状态观测数据，具体特征包括：动作（action）和观测状态（observation.state），均为6维浮点数组，表示肩部平移、肩部升降、肘部弯曲、手腕弯曲、手腕旋转和夹爪位置；观测图像包括顶部摄像头（observation.images.top）和手腕摄像头（observation.images.wrist）的视频，分辨率均为480x640，3通道，帧率30fps；以及触觉加速度计谱图（observation.images.tactile_spectrogram_accelero_10kHz_nfft_512），分辨率为224x224，3通道。数据集还包含时间戳、帧索引、episode索引等元数据。总共有60个episodes，49437帧，数据以parquet文件格式存储，视频以mp4格式存储，总数据大小约300MB。机器人类型为so101_follower_dragon_tactile_bench，主要用于训练任务。This dataset is a robotics dataset created using LeRobot. It contains robot action and state observation data, with specific features including: action and observation.state, both as 6-dimensional float arrays representing shoulder translation, shoulder elevation, elbow flexion, wrist flexion, wrist rotation, and gripper position; observation images include videos from a top camera (observation.images.top) and a wrist camera (observation.images.wrist), both with a resolution of 480x640, 3 channels, and a frame rate of 30fps; as well as tactile accelerometer spectrograms (observation.images.tactile_spectrogram_accelero_10kHz_nfft_512) with a resolution of 224x224 and 3 channels. The dataset also includes metadata such as timestamps, frame indices, and episode indices. There are a total of 60 episodes and 49437 frames, with data stored in parquet file format and videos in mp4 format, totaling approximately 300MB in size. The robot type is so101_follower_dragon_tactile_bench, and it is primarily used for training tasks.