Adversarial Examples Detection with Enhanced Image Difference Features based on Local Histogram Equalization

TL;DR

This paper introduces a detection framework for adversarial examples that enhances high-frequency image features through local histogram equalization, improving detection accuracy without altering existing models.

Contribution

It proposes a novel high-frequency feature enhancement method that can be integrated with existing detection models to improve adversarial example detection.

Findings

Enhanced detection accuracy on adversarial examples

Reduced deployment cost for detection systems

Compatibility with existing detection models

Abstract

Deep Neural Networks (DNNs) have recently made significant progress in many fields. However, studies have shown that DNNs are vulnerable to adversarial examples, where imperceptible perturbations can greatly mislead DNNs even if the full underlying model parameters are not accessible. Various defense methods have been proposed, such as feature compression and gradient masking. However, numerous studies have proven that previous methods create detection or defense against certain attacks, which renders the method ineffective in the face of the latest unknown attack methods. The invisibility of adversarial perturbations is one of the evaluation indicators for adversarial example attacks, which also means that the difference in the local correlation of high-frequency information in adversarial examples and normal examples can be used as an effective feature to distinguish the two. Therefore, we propose an adversarial example detection framework based on a high-frequency information enhancement strategy, which can effectively extract and amplify the feature differences between adversarial examples and normal examples. Experimental results show that the feature augmentation module can be combined with existing detection models in a modular way under this framework. Improve the detector's performance and reduce the deployment cost without modifying the existing detection model.