DiFF

该数据集名为DiFF，包含了超过50万张图像，这些图像是通过13种不同的生成方法在四种条件下合成的：文本到图像、图像到图像、人脸交换和人脸编辑。数据集还包括了多种生成方法，如Midjourney和Stable Diffusion。规模上，该数据集拥有超过50万张图像，其任务是对深度伪造进行检测。

This dataset, named DiFF, contains over 500,000 synthetic images generated via 13 distinct generation methods under four conditions: text-to-image, image-to-image, face swapping, and face editing. It covers a variety of generation techniques, including representative tools such as Midjourney and Stable Diffusion. With a scale of over 500,000 images, its core task is deepfake detection.