stevenworkspace/eval_take_mvactd_1
收藏资源简介:
该数据集是一个用于机器人技术的数据集,使用LeRobot工具创建。它包含一个episode,总计983帧,涉及一个任务。数据以parquet格式存储,视频以mp4格式存储,帧率为30fps。数据集特征包括16维的动作和状态向量,分别表示左右机械臂的关节位置(如left_joint_0.pos到right_left_carriage_joint.pos)和速度(x.vel和theta.vel)。观测数据包括三个摄像头的视频:cam_high(高位摄像头)、cam_left_wrist(左手腕摄像头)和cam_right_wrist(右手腕摄像头),每个视频分辨率为480x640像素,3通道(RGB)。此外,还包含时间戳、帧索引、episode索引、索引和任务索引等元数据。该数据集适用于机器人控制、视觉导航和强化学习等研究。
This dataset is a robotics-oriented dataset developed with the LeRobot toolkit. It contains one episode with a total of 983 frames corresponding to a single task. The dataset is stored in Parquet format, while the associated videos are saved in MP4 format with a frame rate of 30 fps. The dataset features 16-dimensional action and state vectors, which respectively represent the joint positions (e.g., left_joint_0.pos to right_left_carriage_joint.pos) and velocities (x.vel and theta.vel) of the left and right robotic arms. The observation data includes videos from three cameras: cam_high (top-mounted camera), cam_left_wrist (left wrist camera), and cam_right_wrist (right wrist camera). Each video has a resolution of 480×640 pixels with 3 RGB channels. Additionally, it includes metadata such as timestamps, frame indices, episode indices, indices, and task indices. This dataset is suitable for research in fields such as robotic control, visual navigation, and reinforcement learning.




