Exploring Disentangled Content Information for Face Forgery Detection
This addresses the issue of overfitting in face forgery detection for security applications, but it is incremental as it builds on existing methods with novel constraints.
The paper tackles the problem of face forgery detectors overfitting to dataset biases by focusing on content information instead of artifact traces, and proposes a disentanglement framework with constraints to remove content information and enhance feature independence, achieving competitive performance.
Convolutional neural network based face forgery detection methods have achieved remarkable results during training, but struggled to maintain comparable performance during testing. We observe that the detector is prone to focus more on content information than artifact traces, suggesting that the detector is sensitive to the intrinsic bias of the dataset, which leads to severe overfitting. Motivated by this key observation, we design an easily embeddable disentanglement framework for content information removal, and further propose a Content Consistency Constraint (C2C) and a Global Representation Contrastive Constraint (GRCC) to enhance the independence of disentangled features. Furthermore, we cleverly construct two unbalanced datasets to investigate the impact of the content bias. Extensive visualizations and experiments demonstrate that our framework can not only ignore the interference of content information, but also guide the detector to mine suspicious artifact traces and achieve competitive performance.