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

Face recognition (FR) has recently made substantial progress and achievedhigh accuracy on standard benchmarks. However, it has raised security concernsin enormous FR applications because deep CNNs are unusually vulnerable toadversarial examples, and it is still lack of a comprehensive robustnessevaluation before a FR model is deployed in safety-critical scenarios. Tofacilitate a better understanding of the adversarial vulnerability on FR, wedevelop an adversarial robustness evaluation library on FR named\textbf{RobFR}, which serves as a reference for evaluating the robustness ofdownstream tasks. Specifically, RobFR involves 15 popular naturally trained FRmodels, 9 models with representative defense mechanisms and 2 commercial FR APIservices, to perform the robustness evaluation by using various adversarialattacks as an important surrogate. The evaluations are conducted under diverseadversarial settings in terms of dodging and impersonation, $\ell_2$ and$\ell_\infty$, as well as white-box and black-box attacks. We further propose alandmark-guided cutout (LGC) attack method to improve the transferability ofadversarial examples for black-box attacks by considering the specialcharacteristics of FR. Based on large-scale evaluations, the commercial FR APIservices fail to exhibit acceptable performance on robustness evaluation, andwe also draw several important conclusions for understanding the adversarialrobustness of FR models and providing insights for the design of robust FRmodels. RobFR is open-source and maintains all extendable modules, i.e.,\emph{Datasets}, \emph{FR Models}, \emph{Attacks\&Defenses}, and\emph{Evaluations} at\url{https://github.com/ShawnXYang/Face-Robustness-Benchmark}, which will becontinuously updated to promote future research on robust FR.

RobFR: Benchmarking Adversarial Robustness on Face Recognition