gsethia08/part3-insert_remove_cups_from_rack

该数据集是一个用于机器人操作的视觉-动作数据集，由LeRobot创建。它专注于手部机器人任务，包含6个不同的任务，总共有1773个episodes和431206帧数据。数据以视频和状态动作对的形式存储，视频分辨率为1080x1920，帧率为30fps。特征包括来自ego视角的相机观测、机器人状态（如左右手腕和手指的坐标）、相机内外参数以及动作向量。数据集适用于机器人学习研究，特别是涉及视觉输入和精细动作控制的任务。许可证为Apache-2.0。This dataset is a vision-action dataset for robotic manipulation, created by LeRobot. It focuses on hand robot tasks, comprising 6 distinct tasks with a total of 1773 episodes and 431206 frames. The data is stored in video and state-action pairs, with video resolution of 1080x1920 and a frame rate of 30fps. Features include ego-centric camera observations, robot states (e.g., coordinates of left and right wrists and fingers), camera intrinsic and extrinsic parameters, and action vectors. The dataset is suitable for robotics learning research, particularly tasks involving visual input and fine-grained motion control. The license is Apache-2.0.