English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep Neural Networks (DNNs) have recently made significant progress in manyfields. However, studies have shown that DNNs are vulnerable to adversarialexamples, where imperceptible perturbations can greatly mislead DNNs even ifthe full underlying model parameters are not accessible. Various defensemethods have been proposed, such as feature compression and gradient masking.However, numerous studies have proven that previous methods create detection ordefense against certain attacks, which renders the method ineffective in theface of the latest unknown attack methods. The invisibility of adversarialperturbations is one of the evaluation indicators for adversarial exampleattacks, which also means that the difference in the local correlation ofhigh-frequency information in adversarial examples and normal examples can beused as an effective feature to distinguish the two. Therefore, we propose anadversarial example detection framework based on a high-frequency informationenhancement strategy, which can effectively extract and amplify the featuredifferences between adversarial examples and normal examples. Experimentalresults show that the feature augmentation module can be combined with existingdetection models in a modular way under this framework. Improve the detector'sperformance and reduce the deployment cost without modifying the existingdetection model.

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