CycleGAN without checkerboard artifacts for counter-forensics of fake-image detection
This work is significant for researchers in digital forensics and image manipulation, as it demonstrates a method to bypass common fake-image detection techniques, posing a challenge for current forensic tools.
This paper addresses the problem of counter-forensics against fake-image detection by proposing a novel CycleGAN that generates images without checkerboard artifacts. This allows the generated fake images to evade detection methods that rely on the presence of these artifacts.
In this paper, we propose a novel CycleGAN without checkerboard artifacts for counter-forensics of fake-image detection. Recent rapid advances in image manipulation tools and deep image synthesis techniques, such as Generative Adversarial Networks (GANs) have easily generated fake images, so detecting manipulated images has become an urgent issue. Most state-of-the-art forgery detection methods assume that images include checkerboard artifacts which are generated by using DNNs. Accordingly, we propose a novel CycleGAN without any checkerboard artifacts for counter-forensics of fake-mage detection methods for the first time, as an example of GANs without checkerboard artifacts.