trash-in-trashcan-7-24-26

该数据集是一个用于机器人学的数据集，使用LeRobot工具创建。它包含了机器人执行任务时的动作和观测数据，具体特征包括：动作（11维浮点数组，如速度、接触向量等）、观测状态（44维浮点数组，涉及机器人位置、旋转、传感器读数如激光测距、压力、张力等）、多个摄像头视频（包括夹爪摄像头、俯视摄像头和两个锚定摄像头，视频分辨率和帧率各异）、时间戳、帧索引、剧集索引、索引和任务索引。数据集总共有21个剧集、21624帧、1个任务，数据文件大小为100MB，视频文件大小为200MB，机器人类型为stringman。数据以Parquet格式存储，适用于机器人控制和机器学习研究。This dataset is designed for robotics and was created using the LeRobot tool. It includes action and observation data from robot task execution, with specific features such as: actions (11-dimensional floating-point arrays, e.g., velocity, contact vectors), observation states (44-dimensional floating-point arrays, covering robot position, rotation, and sensor readings like laser ranging, pressure, tension), multiple camera videos (including gripper camera, top-view camera, and two anchored cameras, with varying resolutions and frame rates), timestamps, frame index, episode index, index, and task index. The dataset comprises 21 episodes, 21624 frames, and 1 task, with a data file size of 100MB and video file size of 200MB, and the robot type is stringman. The data is stored in Parquet format and is suitable for robotics control and machine learning research.