zigg-ai/content-regions-distrib-yolo-10k

这个数据集包含了7个类别的YOLO格式内容标注，分别是：数字图形、数字演示、数字屏幕录制、数字视频游戏、多讲者视频、场景视频和访谈视频。数据集分为训练集和验证集，其中训练集有7164张图像，验证集有1792张图像。数据集遵循YOLO格式，目录结构包括图像目录、标签目录和配置文件。每个标签文件包含的注释格式为：类别ID、图像中心点X坐标、图像中心点Y坐标、宽度、高度。

This dataset contains 7 classes of content annotations in YOLO format: digital_graphics, digital_presentation, digital_screencast, digital_videogame, footage_multi_speaker, footage_scene, and footage_talkinghead. The dataset is divided into a training set and a validation set, with 7164 images in the training set and 1792 images in the validation set. The dataset follows the YOLO format with a directory structure including an images directory, a labels directory, and a configuration file. Each label file contains annotations in the format: class_id, x_center, y_center, width, height.