YOLO2431/clean_up_table_v2_plus

该数据集是一个机器人学数据集，专门用于双手机器人（机器人类型为yam_bimanual）的控制或模仿学习任务。它包含300个总集数，246897个总帧数，覆盖4个不同任务，并包含900个总视频。数据以30 FPS的帧率采集，存储为Parquet格式文件。数据集的结构包括多个特征：观察状态（32维浮点数组，编码左右手的位置、6D旋转、夹爪状态和关节角度）、动作（类似32维数组，表示控制命令）、以及来自多个摄像头的视频观察（如头部摄像头、左腕摄像头和右腕摄像头，每个视频分辨率为720x1280，3通道，H.264编码）。此外，还包含时间戳、帧索引、集索引、索引和任务索引等元数据。数据集使用Apache 2.0许可证，并基于LeRobot代码库（版本v2.1）创建，适用于训练和评估机器人学习模型。This dataset is a robotics dataset designed for control or imitation learning tasks with a bimanual robot (robot type: yam_bimanual). It includes 300 total episodes, 246897 total frames, covering 4 distinct tasks, and contains 900 total videos. Data is captured at 30 FPS and stored in Parquet format files. The dataset structure comprises multiple features: observation state (a 32-dimensional float32 array encoding left and right hand positions, 6D rotations, gripper states, and joint angles), action (a similar 32-dimensional array representing control commands), and video observations from multiple cameras (e.g., head camera, left wrist camera, and right wrist camera, each with a resolution of 720x1280, 3 channels, H.264 encoded). Additionally, metadata such as timestamps, frame indices, episode indices, indices, and task indices are included. The dataset uses the Apache 2.0 license and is created based on the LeRobot codebase (version v2.1), suitable for training and evaluating robot learning models.