Tsagkas/dataset-feat-v2

该数据集是通过LeRobot项目创建的，主要用于机器人技术领域。数据集包含10个episodes，共1681帧，数据文件大小为100MB，视频文件大小为200MB，帧率为30fps。数据集的结构包括动作（action）、观测状态（observation.state）、观测图像（observation.images.wrist）、VFM特征（observation.vfm_features）等多个特征。动作和观测状态的数据类型为float32，形状为7维；观测图像的数据类型为视频，分辨率为480x640，3通道；VFM特征的数据类型为float32，形状为16x16x768。数据集还包含时间戳、帧索引、episode索引等辅助信息。

This dataset was created using the LeRobot project and is primarily used in the field of robotics. It contains 10 episodes with a total of 1681 frames. The data files are 100MB in size, and the video files are 200MB in size, with a frame rate of 30fps. The dataset structure includes multiple features such as action, observation.state, observation.images.wrist, observation.vfm_features, etc. The data types for action and observation.state are float32 with a shape of 7 dimensions; the observation.images.wrist is of video type with a resolution of 480x640 and 3 channels; the observation.vfm_features is of float32 type with a shape of 16x16x768. The dataset also includes auxiliary information such as timestamps, frame indices, episode indices, etc.