xiaochyVera/eval_pick_red_mug_dp_robot40_426

该数据集使用LeRobot创建，是一个机器人学数据集，专门用于Droid机器人。它包含30个episodes，总计10553帧，覆盖1个任务，并包含90个视频。数据以parquet格式存储，帧率为15 fps。数据集包含多个特征，如动作（8维关节控制）、观测状态（8维关节状态）、右末端执行器位姿（10维，包括6D旋转和3D平移）、来自三个摄像头（左侧Azure Kinect、前部Azure Kinect和腕部摄像头）的RGB图像（分辨率720x1280），以及时间戳、帧索引、episode索引等元数据。所有数据均用于训练分割，适用于机器人控制、模仿学习或强化学习任务。

This dataset was created using LeRobot and is a robotics dataset specifically for the Droid robot. It contains 30 episodes, totaling 10553 frames, covering 1 task, and includes 90 videos. The data is stored in parquet format with a frame rate of 15 fps. The dataset features multiple components such as actions (8-dimensional joint control), observation states (8-dimensional joint states), right end-effector poses (10-dimensional, including 6D rotation and 3D translation), RGB images from three cameras (left Azure Kinect, front Azure Kinect, and wrist camera) with a resolution of 720x1280, and metadata like timestamps, frame indices, and episode indices. All data is allocated for the training split and is suitable for robotics control, imitation learning, or reinforcement learning tasks.