ases200q2/Aibot2_combined_pick_object_datasets_updatad_2026_05_20_15fps

这是一个用于机器人学习的控制数据集，由LeRobot创建。数据集专门针对Aibot2机器人设计，专注于拾取物体的任务。它包含465个完整的情节（episodes），总计114,771帧数据，帧率为15 fps。数据以parquet格式存储，总数据文件大小为100 MB，视频文件大小为200 MB。数据集包含多种特征：动作数据（包括左右机械臂的位置和四元数姿态，以及左右夹爪的位置）、观测状态数据（包括左右机械臂各7个关节的位置，以及左右夹爪关节的位置）、来自四个摄像头（前、左、右、顶部）的图像观测（每个视频分辨率为240x320，3通道，25 fps，AV1编码），以及时间戳、帧索引、情节索引等元数据。数据集仅包含训练分割（0:465），适用于机器人控制、模仿学习或强化学习的研究和应用。This is a control dataset for robot learning, created using LeRobot. The dataset is specifically designed for the Aibot2 robot and focuses on the task of picking objects. It contains 465 total episodes with 114,771 total frames at 15 fps. The data is stored in parquet format, with a total data file size of 100 MB and video file size of 200 MB. The dataset includes various features: action data (including left and right arm position and quaternion pose, and left and right gripper positions), observation state data (including positions of 7 joints each for left and right arms, and left and right gripper joints), image observations from four cameras (front, left, right, top; each video at 240x320 resolution, 3 channels, 25 fps, AV1 codec), and metadata such as timestamp, frame index, episode index. The dataset only includes a train split (0:465) and is suitable for research and applications in robot control, imitation learning, or reinforcement learning.