Patch-Discontinuity Mining for Generalized Deepfake Detection
This addresses the threat of realistic fake facial images to personal privacy and online integrity, representing an incremental improvement over existing methods.
The paper tackles the problem of deepfake detection by proposing GenDF, a framework that transfers a large-scale vision model to capture discriminative patterns and enhance generalization, achieving state-of-the-art performance in cross-domain settings with only 0.28M trainable parameters.
The rapid advancement of generative artificial intelligence has enabled the creation of highly realistic fake facial images, posing serious threats to personal privacy and the integrity of online information. Existing deepfake detection methods often rely on handcrafted forensic cues and complex architectures, achieving strong performance in intra-domain settings but suffering significant degradation when confronted with unseen forgery patterns. In this paper, we propose GenDF, a simple yet effective framework that transfers a powerful large-scale vision model to the deepfake detection task with a compact and neat network design. GenDF incorporates deepfake-specific representation learning to capture discriminative patterns between real and fake facial images, feature space redistribution to mitigate distribution mismatch, and a classification-invariant feature augmentation strategy to enhance generalization without introducing additional trainable parameters. Extensive experiments demonstrate that GenDF achieves state-of-the-art generalization performance in cross-domain and cross-manipulation settings while requiring only 0.28M trainable parameters, validating the effectiveness and efficiency of the proposed framework.