A New Approach to Improve Learning-based Deepfake Detection in Realistic Conditions
This work addresses robustness in deepfake detection for security applications, but it is incremental as it builds on existing data augmentation methods.
The paper tackles the problem of deepfake detection under realistic conditions by proposing a data augmentation scheme based on real-world image degradation, which improves generalization to unpredictable distortions and unseen datasets.
Deep convolutional neural networks have achieved exceptional results on multiple detection and recognition tasks. However, the performance of such detectors are often evaluated in public benchmarks under constrained and non-realistic situations. The impact of conventional distortions and processing operations found in imaging workflows such as compression, noise, and enhancement are not sufficiently studied. Currently, only a few researches have been done to improve the detector robustness to unseen perturbations. This paper proposes a more effective data augmentation scheme based on real-world image degradation process. This novel technique is deployed for deepfake detection tasks and has been evaluated by a more realistic assessment framework. Extensive experiments show that the proposed data augmentation scheme improves generalization ability to unpredictable data distortions and unseen datasets.