yinongh/automate_task_2
收藏资源简介:
该数据集使用LeRobot创建,专注于机器人任务。数据集包含18个训练集,总帧数为1122,帧率为30fps,数据以parquet格式存储。数据集结构包括动作数据(如9维浮点数组,表示位置和旋转变化)、观测数据(如末端执行器内部力、姿态、深度图像、点云、相机外参、RGB和深度视频等),以及时间戳、帧索引、集索引等元数据。特征涵盖多传感器输入,例如手腕相机和辅助相机的RGB与深度图像,用于支持机器人学习和控制任务。
This dataset was created using LeRobot and focuses on robotics tasks. It contains 18 training episodes with a total of 1122 frames at 30 fps, stored in parquet format. The dataset structure includes action data (e.g., a 9-dimensional float array representing positional and rotational changes), observation data (such as end-effector internal forces, poses, depth images, point clouds, camera extrinsics, RGB and depth videos), and metadata like timestamps, frame indices, and episode indices. Features encompass multi-sensor inputs, including RGB and depth images from wrist and auxiliary cameras, to support robot learning and control tasks.




