RoboCOIN/AI2_Alphabot_2_shake_test_tube_1
收藏资源简介:
该数据集名为AI2_Alphabot_2_shake_test_tube_1,是一个用于机器人操作的现实世界数据集,专注于一个具体的操作任务:从试管架上拿起透明试管,摇晃它,然后将其放回试管架。数据集基于LeRobot格式并完全兼容,包含399个完整操作片段(episodes),总计313,739帧视频数据,大小为24.50 GB。数据采集自名为AI2_Alphabot_2的双臂机器人,配备双指末端执行器,并通过VR控制器进行遥操作。场景设置为零售超市环境,涉及的对象包括透明的试管架和试管。数据集提供了多视角的RGB视频观测,来自四个摄像头:胸前前置摄像头、头部前置摄像头、左腕摄像头和右腕摄像头,分辨率均为640x480,帧率为30 FPS。此外,数据集包含机器人的状态信息(如关节位置、末端执行器位姿)和动作指令(34维),以及丰富的注释信息(如末端执行器速度、加速度、夹爪活动等)。数据集主要用于机器人学习研究,特别是模仿学习、强化学习和操作技能学习。
This dataset, named AI2_Alphabot_2_shake_test_tube_1, is a real-world robotic manipulation dataset focused on a specific task: picking up a transparent test tube from a test tube rack, shaking it, and then placing it back into the rack. It is based on and fully compatible with the LeRobot format, containing 399 complete episodes, totaling 313,739 frames of video data, with a size of 24.50 GB. The data was collected using a dual-arm robot named AI2_Alphabot_2, equipped with a two-finger end-effector and teleoperated via a VR controller. The scene is set in a retail supermarket environment, involving objects such as a transparent test tube rack and a test tube. The dataset provides multi-view RGB video observations from four cameras: front chest RGB, front head RGB, left wrist RGB, and right wrist RGB, all with a resolution of 640x480 and a frame rate of 30 FPS. Additionally, the dataset includes robot state information (e.g., joint positions, end-effector poses) and action commands (34-dimensional), as well as rich annotations (e.g., end-effector velocity, acceleration, gripper activity). The dataset is primarily intended for robotics learning research, particularly imitation learning, reinforcement learning, and manipulation skill learning.




