stevenworkspace/eval_take_act_6
收藏资源简介:
该数据集是通过LeRobot平台创建的机器人数据集,专用于机器人学习任务。数据集包含一个完整的episode,共计1044帧,帧率为30fps,数据以parquet文件格式存储,并附带相应的mp4视频文件。数据集记录了mobileai_robot机器人的动作和观测数据:动作特征包括16个浮点数值,表示左右机械臂的关节位置和速度;观测状态特征同样包含16个浮点数值,反映机器人的实时状态。此外,数据集提供了三个摄像头的图像观测:一个高位摄像头(cam_high)和两个腕部摄像头(cam_left_wrist、cam_right_wrist),所有图像均为480x640分辨率的RGB视频。其他元数据如时间戳、帧索引、episode索引等也包含在内,适用于机器人控制、模仿学习或强化学习等研究。
This dataset is a robotics dataset created using the LeRobot platform, designed for robot learning tasks. It consists of a single episode with 1044 frames at 30fps, stored in parquet format along with accompanying mp4 video files. The dataset captures action and observation data from a mobileai_robot: the action feature includes 16 float values representing joint positions and velocities for both left and right arms; the observation state feature similarly contains 16 float values reflecting the robots real-time status. Additionally, the dataset provides image observations from three cameras: a high camera (cam_high) and two wrist cameras (cam_left_wrist, cam_right_wrist), all with 480x640 resolution RGB video. Metadata such as timestamp, frame index, episode index, and others are included, making it suitable for research in robot control, imitation learning, or reinforcement learning.




