cell1_20260520_laundry_folding_20260520_132519

这是一个机器人学数据集，专为双机械臂操作任务设计，具体聚焦于衣物折叠（laundry folding）场景。数据集使用LeRobot工具创建，包含2个完整的情节（episodes），总计1610帧数据，帧率为30fps。数据内容涵盖机械臂的动作（action）信息，如位置、旋转和夹爪宽度，以及状态观测（observation），包括编码器角度、IMU加速度和陀螺仪数据、位姿（位置和四元数）和夹爪状态。此外，数据集提供了多视角图像数据，来自左腕和右腕的RGB和深度摄像头，以及基座的两个RGB摄像头，图像分辨率分别为480x640和768x1024。数据以parquet和视频格式存储，适用于机器人控制、模仿学习或强化学习研究。This is a robotics dataset tailored for dual-arm manipulation tasks, with a specific focus on laundry folding scenarios. Created using the LeRobot toolkit, the dataset includes 2 full episodes, totaling 1610 frames of data at 30fps. The data encompasses robotic arm action information, including position, rotation, and gripper width, as well as state observations consisting of encoder angles, IMU acceleration and gyroscope measurements, poses (position and quaternion), and gripper states. Additionally, the dataset provides multi-view image data from RGB and depth cameras on the left and right wrists, plus two RGB cameras mounted on the robot base, with image resolutions of 480x640 and 768x1024 respectively. The data is stored in Parquet and video formats, and is suitable for research in robotic control, imitation learning, and reinforcement learning.