mug_in_bin_20250821_subsampled_noWrist_heatmapGoal

这是一个机器人控制数据集，使用LeRobot创建。包含20个训练片段，总计4926帧数据和60个视频，帧率为15fps。数据集采集自双机械臂系统，包含丰富的传感器数据：左右机械臂的18维关节状态和动作控制信号、末端执行器的10维位姿信息、Azure Kinect相机的720p彩色图像和深度图像、机械臂夹爪的点云数据（500个点）以及时间戳和帧索引等信息。数据以parquet格式存储，适用于机器人学习和控制算法研究。

This is a robotic control dataset created using LeRobot. It includes 20 training episodes, totaling 4926 frames across 60 videos with a frame rate of 15 fps. The dataset was collected from a dual-arm robotic system, and contains rich sensor data: 18-dimensional joint states and action control signals for both the left and right robotic arms, 10-dimensional pose information of the end-effectors, 720p color and depth images captured by Azure Kinect cameras, 500-point point cloud data from the robotic grippers, as well as timestamps and frame indices. The data is stored in Parquet format, and is suitable for research in robotic learning and control algorithm studies.