juyoungggg/0509_setting_1st

该数据集是一个机器人领域的数据集，专门用于双机械臂跟随任务（bi_so_follower）。它包含100个训练episodes，总计112,428帧数据，覆盖1个任务。数据集以parquet格式存储数据文件，以mp4格式存储视频文件，视频帧率为30fps。特征包括：12维的动作向量（对应左右机械臂的12个关节位置，如肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置）和12维的观测状态向量（与动作相同）。图像观测来自四个摄像头视角：左腕部摄像头、左后上方摄像头、右腕部摄像头和右前上方摄像头，每个图像分辨率为480x640，包含3个通道（RGB），视频编码为av1。此外，数据集还包含时间戳、帧索引、episode索引、任务索引等元数据。所有数据仅用于训练分割（train: 0:100）。数据集使用Apache 2.0许可证，并通过LeRobot工具创建。This dataset is a robotics dataset specifically designed for a dual-arm following task (bi_so_follower). It contains 100 training episodes, totaling 112,428 frames, and covers 1 task. The data is stored in parquet format for data files and mp4 format for video files, with a video frame rate of 30fps. Features include: a 12-dimensional action vector (corresponding to the positions of 12 joints on the left and right robotic arms, such as shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper positions) and a 12-dimensional observation state vector (identical to the action). Image observations are captured from four camera perspectives: left wrist camera, left top-back camera, right wrist camera, and right top-front camera, each with a resolution of 480x640 and 3 channels (RGB), encoded in av1. Additionally, the dataset includes metadata such as timestamps, frame indices, episode indices, and task indices. All data is allocated to the training split (train: 0:100). The dataset is licensed under Apache 2.0 and was created using the LeRobot tool.