cell1_20260524_mohamed-salah_grocery_bagging_cycle_20260524_111558

该数据集是一个机器人操作数据集，使用LeRobot框架创建，专门用于机器人学习任务。数据集记录了starpilot_yam_gripper双机械臂系统执行杂货装袋任务的操作过程。它包含5个完整的训练片段（episodes），总计2051帧数据，帧率为30fps。数据内容主要包括：1) 动作数据：14维浮点向量，控制两个机械臂的末端执行器位置（x, y, z）、旋转（rx, ry, rz）和抓取宽度。2) 状态观测数据：32维浮点向量，包括两个机械臂的关节编码器角度、IMU加速度和角速度、末端执行器位姿（位置和四元数）、以及夹爪开合状态和距离。3) 多视角视觉数据：来自左腕、右腕和基座摄像头的RGB图像和深度图像视频，分辨率分别为480x640（腕部摄像头）和768x1024（基座摄像头）。所有数据以Parquet格式存储，总大小约300MB（数据文件100MB，视频文件200MB）。该数据集适用于机器人模仿学习、强化学习等研究。

This is a robotic manipulation dataset created with the LeRobot framework, specifically designed for robot learning research. It records the manipulation processes of the starpilot_yam_gripper dual-arm robotic system during the grocery bagging task. The dataset contains 5 complete training episodes, totaling 2051 frames of data with a frame rate of 30fps.

The dataset primarily includes three categories of data:
1. Action data: A 14-dimensional floating-point vector that controls the end-effector position (x, y, z), rotation (rx, ry, rz), and gripper opening width of the two robotic arms.
2. State observation data: A 32-dimensional floating-point vector covering joint encoder angles of the two robotic arms, IMU acceleration and angular velocity, end-effector pose (position and quaternion), as well as gripper opening status and distance.
3. Multi-view visual data: RGB and depth image videos captured by left wrist, right wrist, and base-mounted cameras, with resolutions of 480x640 for wrist cameras and 768x1024 for the base camera respectively.

All data is stored in Parquet format, with a total size of approximately 300MB (100MB for data files and 200MB for video files). This dataset is applicable to research in robotic imitation learning, reinforcement learning, and other related robot learning fields.