lesc-unifi/dragon

DRAGON是一个大规模的数据集，包含由25种扩散模型生成的250万张训练图像和10万张测试图像。这个数据集旨在支持新型多媒体取证工具的开发，特别是用于合成图像检测和模型归因任务。数据集覆盖了多种主题，并提供不同大小的版本以满足不同的研究需求。

DRAGON is a large-scale dataset consisting of 2.5 million training images and 100,000 test images generated by 25 diffusion models. It is designed to support the development of new multimedia forensics tools, specifically for synthetic image detection and model attribution tasks. The dataset covers a diverse range of subjects and is available in various sizes to cater to different research needs.