pyrrosk/aloha_pipes_single_easy_grasp
收藏资源简介:
该数据集是一个用于机器人学习的仿真数据集,使用LeRobot平台创建。数据集包含ALOHA类型机器人的观察和动作数据,具体包括:状态观察(14维浮点数,表示左右机械臂的关节位置,如腰部、肩部、肘部等)、动作(14维浮点数,与状态观察对应)、三个摄像头的图像数据(顶部摄像头、左腕摄像头和右腕摄像头,均为224x224分辨率的RGB视频,帧率50fps),以及时间戳、帧索引、episode索引等元数据。数据集总共有51个episodes,32588帧,所有数据均用于训练,文件格式为Parquet,总数据大小约100MB,视频文件大小约200MB。该数据集适用于机器人控制、模仿学习等任务。
This dataset is a simulation dataset for robotics learning, created using the LeRobot platform. It contains observation and action data for an ALOHA-type robot, including: state observations (14-dimensional float values representing joint positions of left and right robotic arms, such as waist, shoulder, elbow, etc.), actions (14-dimensional float values corresponding to state observations), image data from three cameras (top camera, left wrist camera, and right wrist camera, all as RGB videos with 224x224 resolution at 50fps), and metadata such as timestamp, frame index, and episode index. The dataset comprises 51 episodes and 32,588 frames, all used for training. The data is stored in Parquet format, with a total data size of approximately 100MB and video files of about 200MB. It is suitable for tasks like robot control and imitation learning.




