arminfg/op2reader3_duplicated

该数据集使用LeRobot创建，包含17个剧集、7679个帧和34个视频。数据集的结构包括动作、观测状态、前视相机图像、顶视相机图像等多种特征。数据集在Apache-2.0许可证下发布。数据被分为训练集和测试集，训练集包括从第0个到第17个剧集。数据集包含肩部旋转位置、肩部提升位置、肘部弯曲位置、手腕弯曲位置、手腕滚动位置和夹爪位置等特征，以及前视和顶视相机的图像。数据集没有音频，并使用av1编解码器进行视频压缩。

This dataset was created using LeRobot and contains 17 episodes, 7679 frames, and 34 videos. It is structured with features such as action, observation.state, front camera images, and top camera images. The dataset is released under the Apache-2.0 license. The data is split into a training set and a test set, with the training set ranging from episode 0 to 17. The dataset includes features like shoulder_pan.position, shoulder_lift.position, elbow_flex.position, wrist_flex.position, wrist_roll.position, and gripper.position, along with front and top camera images. The dataset does not have audio and uses the av1 codec for video compression.