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

Recent years have witnessed unprecedented success achieved by deep learningmodels in the field of computer vision. However, their vulnerability towardscarefully crafted adversarial examples has also attracted the increasingattention of researchers. Motivated by the observation that adversarialexamples are due to the non-robust feature learned from the original dataset bymodels, we propose the concepts of salient feature(SF) and trivial feature(TF).The former represents the class-related feature, while the latter is usuallyadopted to mislead the model. We extract these two features with coupledgenerative adversarial network model and put forward a novel detection anddefense method named salient feature extractor (SFE) to defend againstadversarial attacks. Concretely, detection is realized by separating andcomparing the difference between SF and TF of the input. At the same time,correct labels are obtained by re-identifying SF to reach the purpose ofdefense. Extensive experiments are carried out on MNIST, CIFAR-10, and ImageNetdatasets where SFE shows state-of-the-art results in effectiveness andefficiency compared with baselines. Furthermore, we provide an interpretableunderstanding of the defense and detection process.

Salient Feature Extractor for Adversarial Defense on Deep Neural Networks