XAI-Based Detection of Adversarial Attacks on Deepfake Detectors

View on arXiv ← Back to list

Authors: Ben Pinhasov, Raz Lapid, Rony Ohayon, Moshe Sipper, Yehudit Aperstein

Published: 2024-03-05 13:25:30+00:00

AI Summary

This paper proposes an XAI-based method for detecting adversarial attacks on deepfake detectors. It uses XAI to generate interpretability maps, which are then processed along with the original image by a pretrained feature extractor to train a classifier that detects adversarial attacks without affecting the deepfake detector's performance.

Abstract

We introduce a novel methodology for identifying adversarial attacks on deepfake detectors using eXplainable Artificial Intelligence (XAI). In an era characterized by digital advancement, deepfakes have emerged as a potent tool, creating a demand for efficient detection systems. However, these systems are frequently targeted by adversarial attacks that inhibit their performance. We address this gap, developing a defensible deepfake detector by leveraging the power of XAI. The proposed methodology uses XAI to generate interpretability maps for a given method, providing explicit visualizations of decision-making factors within the AI models. We subsequently employ a pretrained feature extractor that processes both the input image and its corresponding XAI image. The feature embeddings extracted from this process are then used for training a simple yet effective classifier. Our approach contributes not only to the detection of deepfakes but also enhances the understanding of possible adversarial attacks, pinpointing potential vulnerabilities. Furthermore, this approach does not change the performance of the deepfake detector. The paper demonstrates promising results suggesting a potential pathway for future deepfake detection mechanisms. We believe this study will serve as a valuable contribution to the community, sparking much-needed discourse on safeguarding deepfake detectors.

Key findings

The proposed method effectively detects adversarial attacks on deepfake detectors with high accuracy, particularly using the Saliency XAI method. The approach does not negatively impact the performance of the underlying deepfake detector. Generalization across different deepfake detectors was also demonstrated, although with reduced accuracy compared to using the same detector for both training and testing.

Approach

The approach leverages XAI to generate interpretability maps of a deepfake detector's decision-making process. These maps, along with the input image, are fed into a pretrained feature extractor and then used to train a classifier to identify adversarial attacks.

Datasets

FaceForensics++ (FF++) dataset

Model(s)

XceptionNet, EfficientNetB4ST, ResNet50 (as a backbone for the adversarial attack detector)

Author countries

Israel

← Previous