What's wrong with this video? Comparing Explainers for Deepfake Detection
This work addresses the need for explainability in deepfake detection to enhance trust and usability, but it is incremental as it adapts existing methods to a specific domain.
The paper tackled the problem of understanding why deepfake detectors label videos as fake by developing and comparing white-box, black-box, and model-specific explanation techniques, such as SHAP, GradCAM, and self-attention models, applied to EfficientNet-based detectors on the DFDC dataset, and proposed metrics and conducted a user survey to evaluate their usefulness.
Deepfakes are computer manipulated videos where the face of an individual has been replaced with that of another. Software for creating such forgeries is easy to use and ever more popular, causing serious threats to personal reputation and public security. The quality of classifiers for detecting deepfakes has improved with the releasing of ever larger datasets, but the understanding of why a particular video has been labelled as fake has not kept pace. In this work we develop, extend and compare white-box, black-box and model-specific techniques for explaining the labelling of real and fake videos. In particular, we adapt SHAP, GradCAM and self-attention models to the task of explaining the predictions of state-of-the-art detectors based on EfficientNet, trained on the Deepfake Detection Challenge (DFDC) dataset. We compare the obtained explanations, proposing metrics to quantify their visual features and desirable characteristics, and also perform a user survey collecting users' opinions regarding the usefulness of the explainers.