raghav-proprio/assemble-trossen-toy-v3-test
收藏资源简介:
该数据集是一个机器人学数据集,使用LeRobot工具创建,专门针对Trossen WXAI双手机器人。数据集包含2个总集数、963个总帧数和1个总任务数,以30fps的帧率采集。数据以parquet格式存储,视频以mp4格式存储。数据集的特征包括观察状态(如右臂的6个关节位置和右夹爪位置)、动作(与观察状态相同维度)、末端执行器位置和姿态(以四元数表示),以及来自两个摄像头的图像观察:一个高位摄像头(cam_high)和一个右腕摄像头(cam_right_wrist),每个图像分辨率为480x640像素、3个通道。此外,还包括时间戳、帧索引、集索引、索引和任务索引等元数据。该数据集适用于机器人控制、模仿学习或强化学习任务。
This is a robotics dataset developed using the LeRobot toolkit, specifically tailored for the Trossen WXAI dual-arm robot. It contains 2 total episodes, 963 total frames, and 1 total task, with data collected at a frame rate of 30 fps. The structured dataset data is stored in Parquet format, while the accompanying videos are saved in MP4 format. The dataset includes the following features: observation states (e.g., 6 joint positions of the right arm and the right gripper position), actions with the same dimension as the observation states, end-effector positions and orientations represented by quaternions, as well as image observations from two cameras: a top-mounted camera (cam_high) and a right wrist camera (cam_right_wrist). Each image has a resolution of 480×640 pixels and 3 color channels. Additionally, it includes metadata such as timestamps, frame indices, episode indices, sample indices, and task indices. This dataset is suitable for robotics control, imitation learning, or reinforcement learning tasks.





