cell1_20260524_mohamed-salah_grocery_bagging_cycle_20260524_112123

该数据集是一个机器人操作数据集，基于LeRobot框架构建，专门设计用于机器人学习任务。它包含5个完整的情节，总计1846帧，专注于单个任务，例如袋装杂货操作。数据集详细记录了双机械臂系统的动作控制信息，包括位置、旋转和抓取宽度，以及观察状态数据，如编码器角度、IMU（惯性测量单元）数据、位姿和抓取器距离。此外，数据集提供多视角的摄像头视觉数据，包括左腕和右腕的RGB及深度图像，以及基座摄像头的RGB图像，以支持视觉感知。所有数据以parquet文件格式存储，视频部分以mp4格式保存，帧率为30fps，适用于机器人模仿学习或强化学习的研究和应用。This dataset is a robotic manipulation dataset built using the LeRobot framework, specifically designed for robot learning tasks. It contains 5 complete episodes, totaling 1846 frames, and focuses on a single task, such as bagged grocery manipulation. The dataset includes detailed action control information for a dual-arm robotic system, covering position, rotation, and gripper width, as well as observation state data like encoder angles, IMU (Inertial Measurement Unit) data, poses, and gripper distances. Additionally, it provides multi-view camera visual data, including RGB and depth images from the left and right wrists, along with RGB images from a base camera, to support visual perception. All data is stored in parquet file format, with video portions saved in mp4 format at a frame rate of 30fps, making it suitable for research and applications in robot imitation learning or reinforcement learning.