cell1_20260520_laundry_folding_20260520_130712

该数据集使用LeRobot创建，包含机器人执行任务时的多模态数据。数据集涉及机器人类型为starpilot_yam_gripper，总共有2个episodes和1742帧，覆盖1个任务。数据以30fps采集，包括动作数据（14维，如机械臂位置、旋转和夹爪宽度）、状态观测数据（32维，如编码器角度、IMU数据和姿态）以及多个摄像头图像：左腕0号摄像头RGB图像和深度图像（480x640）、右腕0号摄像头RGB图像和深度图像（480x640）、基座0号摄像头RGB图像（480x640）和基座1号摄像头RGB图像（768x1024）。所有图像均为视频格式，使用h264编解码器。数据集结构以parquet文件存储，用于机器人学习和控制任务。This dataset was created using LeRobot and encompasses multimodal data collected during robotic task execution. The target robotic platform is the starpilot_yam_gripper, with a total of 2 episodes and 1742 frames, covering one single task. All data was acquired at 30 frames per second (fps), and includes: 14-dimensional action data (e.g., robotic arm position, rotation, and gripper width), 32-dimensional state observation data (e.g., encoder angles, IMU data, and pose), and multiple camera image streams: RGB and depth images from the left wrist 0 camera (480×640), RGB and depth images from the right wrist 0 camera (480×640), RGB images from the base 0 camera (480×640), and RGB images from the base 1 camera (768×1024). All image data is stored in video format using the H.264 codec. The dataset is structured and stored as Parquet files, and is designed for robotic learning and control applications.