Reality Transform Adversarial Generators for Image Splicing Forgery Detection and Localization

When many forgery images become more and more realistic with help of image editing tools and convolutional neural networks (CNNs), authenticators need to improve their ability to verify these forgery images. The process of generating and detecting forgery images is the same as the principle of Generative Adversarial Networks (GANs). In this paper, since the retouching progress of forgery images requires to suppress the tampering artifacts and to keep the structural information, we consider this retouching progress as an image style transform, and then propose a fake-to-realistic transform generator G T . For detecting the tampered regions, a localization generator G M is proposed too, which is based on a multi-decoder-single-task strategy. By adversarial training two generators, the proposed α-learnable whitening and coloring transform (αlearnable WCT) block in G T automatically suppress the tampering artifacts in the forgery images. Meanwhile, the detection and localization abilities of G M will be improved by learning the forgery images retouched by G T . The experiment results demonstrate that the proposed two generators in GAN can simulate confrontation between the faker and the authenticator well; the localization generator G M outperforms the state-of-the-art methods in splicing forgery detection and localization on four public datasets.