cf-group-4/cloth_folding_final_may19

该数据集是一个用于机器人任务的数据集，具体涉及布料折叠任务。它包含91个训练集，总计148,440帧数据，帧率为30fps。数据集的结构包括动作特征（如肩部、肘部、腕部和夹爪的位置）、观测状态（与动作特征相同）、来自前置摄像头的图像观测（分辨率为480x640，3通道彩色视频）、时间戳、帧索引、集索引和任务索引。数据以Parquet格式存储，视频以MP4格式存储，使用AV1编解码器。该数据集使用LeRobot工具创建，适用于机器人学习和控制研究。

This dataset is designed for robotics tasks, specifically focusing on cloth folding. It includes 91 training episodes with a total of 148,440 frames at 30fps. The dataset structure comprises action features (e.g., positions of shoulder, elbow, wrist, and gripper), observation states (same as action features), image observations from a front camera (480x640 resolution, 3-channel color video), timestamps, frame indices, episode indices, and task indices. Data is stored in Parquet format, with videos in MP4 format using the AV1 codec. Created using the LeRobot tool, it is suitable for robotics learning and control research.