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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Deep neural networks (DNN) have been a de facto standard for nowadaysbiometric recognition solutions. A serious, but still overlooked problem inthese DNN-based recognition systems is their vulnerability against adversarialattacks. Adversarial attacks can easily cause the output of a DNN system togreatly distort with only tiny changes in its input. Such distortions canpotentially lead to an unexpected match between a valid biometric and asynthetic one constructed by a strategic attacker, raising security issue. Inthis work, we show how this issue can be resolved by learning robust biometricfeatures through a deep, information-theoretic framework, which builds upon therecent deep variational information bottleneck method but is carefully adaptedto biometric recognition tasks. Empirical evaluation demonstrates that ourmethod not only offers stronger robustness against adversarial attacks but alsoprovides better recognition performance over state-of-the-art approaches.

A Deep, Information-theoretic Framework for Robust Biometric Recognition