Deepfake Detection without Deepfakes: Generalization via Synthetic Frequency Patterns Injection
This addresses the critical issue of unreliable deepfake detection for security and media integrity, offering a novel approach that is not incremental but a new paradigm for generalization.
The paper tackled the problem of deepfake detectors' limited generalization to unseen generation methods by training them using only pristine images with injected synthetic frequency patterns, achieving state-of-the-art detection and superior generalization across 25 generation methods.
Deepfake detectors are typically trained on large sets of pristine and generated images, resulting in limited generalization capacity; they excel at identifying deepfakes created through methods encountered during training but struggle with those generated by unknown techniques. This paper introduces a learning approach aimed at significantly enhancing the generalization capabilities of deepfake detectors. Our method takes inspiration from the unique "fingerprints" that image generation processes consistently introduce into the frequency domain. These fingerprints manifest as structured and distinctly recognizable frequency patterns. We propose to train detectors using only pristine images injecting in part of them crafted frequency patterns, simulating the effects of various deepfake generation techniques without being specific to any. These synthetic patterns are based on generic shapes, grids, or auras. We evaluated our approach using diverse architectures across 25 different generation methods. The models trained with our approach were able to perform state-of-the-art deepfake detection, demonstrating also superior generalization capabilities in comparison with previous methods. Indeed, they are untied to any specific generation technique and can effectively identify deepfakes regardless of how they were made.