ddduk/Pick_and_place_green_bean_20260526_213911

该数据集是一个机器人学数据集，由LeRobot创建，专注于机器人操作任务。数据集包含25个episodes，总计10907帧，涉及2个不同任务。数据特征包括：动作数据（action），表示机器人关节位置（如肩部平移、肩部升降、肘部弯曲、手腕弯曲、手腕旋转和夹爪位置）；观测状态（observation.state），与动作数据类似，反映机器人当前状态；图像观测，包括顶部摄像头（observation.images.top）和手腕摄像头（observation.images.wrist）采集的视频，分辨率为480x640，帧率为30fps，使用AV1编码；以及时间戳、帧索引、episode索引、任务索引等元数据。机器人类型为so_follower，数据以parquet文件格式存储，视频以mp4格式存储，总数据大小约为100MB，视频文件大小约为200MB。数据集用于训练和评估机器人控制模型，支持从多模态观测中学习动作策略。

This dataset is a robotics dataset created using LeRobot, focusing on robot manipulation tasks. It consists of 25 episodes with a total of 10,907 frames across 2 distinct tasks. The features include: action data representing robot joint positions (e.g., shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper positions); observation state similar to action data, reflecting the current robot state; image observations from a top camera (observation.images.top) and a wrist camera (observation.images.wrist), with video resolution of 480x640, frame rate of 30fps, encoded in AV1; and metadata such as timestamp, frame index, episode index, and task index. The robot type is so_follower. Data is stored in parquet format, with videos in mp4 format, total data size approximately 100MB and video files size approximately 200MB. The dataset is designed for training and evaluating robot control models, enabling learning of action policies from multimodal observations.