Representative Forgery Mining for Fake Face Detection
This work improves fake face detection for security and media verification applications, representing an incremental advancement by enhancing existing methods with a simple augmentation technique.
The paper tackles the problem of fake face detection by addressing detectors' limited focus on specific facial regions, proposing an attention-based data augmentation framework that encourages mining of more representative forgeries, resulting in state-of-the-art performance without structural changes to CNN models.
Although vanilla Convolutional Neural Network (CNN) based detectors can achieve satisfactory performance on fake face detection, we observe that the detectors tend to seek forgeries on a limited region of face, which reveals that the detectors is short of understanding of forgery. Therefore, we propose an attention-based data augmentation framework to guide detector refine and enlarge its attention. Specifically, our method tracks and occludes the Top-N sensitive facial regions, encouraging the detector to mine deeper into the regions ignored before for more representative forgery. Especially, our method is simple-to-use and can be easily integrated with various CNN models. Extensive experiments show that the detector trained with our method is capable to separately point out the representative forgery of fake faces generated by different manipulation techniques, and our method enables a vanilla CNN-based detector to achieve state-of-the-art performance without structure modification.