Detecting GAN-generated Imagery using Color Cues
This addresses the issue of online disinformation and social media manipulation by providing a method to detect synthetic imagery, though it appears incremental as it builds on existing GAN analysis.
The paper tackled the problem of detecting GAN-generated imagery, such as deepfakes, by analyzing color cues in the generating network, and demonstrated effective discrimination between GAN and real camera images.
Image forensics is an increasingly relevant problem, as it can potentially address online disinformation campaigns and mitigate problematic aspects of social media. Of particular interest, given its recent successes, is the detection of imagery produced by Generative Adversarial Networks (GANs), e.g. `deepfakes'. Leveraging large training sets and extensive computing resources, recent work has shown that GANs can be trained to generate synthetic imagery which is (in some ways) indistinguishable from real imagery. We analyze the structure of the generating network of a popular GAN implementation, and show that the network's treatment of color is markedly different from a real camera in two ways. We further show that these two cues can be used to distinguish GAN-generated imagery from camera imagery, demonstrating effective discrimination between GAN imagery and real camera images used to train the GAN.