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

Today, the security of many domains rely on the use of Machine Learning todetect threats, identify vulnerabilities, and safeguard systems from attacks.Recently, transformer architectures have improved the state-of-the-artperformance on a wide range of tasks such as malware detection and networkintrusion detection. But, before abandoning current approaches to transformers,it is crucial to understand their properties and implications on cybersecurityapplications. In this paper, we evaluate the robustness of transformers toadversarial samples for system defenders (i.e., resiliency to adversarialperturbations generated on different types of architectures) and theiradversarial strength for system attackers (i.e., transferability of adversarialsamples generated by transformers to other target models). To that effect, wefirst fine-tune a set of pre-trained transformer, Convolutional Neural Network(CNN), and hybrid (an ensemble of transformer and CNN) models to solvedifferent downstream image-based tasks. Then, we use an attack algorithm tocraft 19,367 adversarial examples on each model for each task. Thetransferability of these adversarial examples is measured by evaluating eachset on other models to determine which models offer more adversarial strength,and consequently, more robustness against these attacks. We find that theadversarial examples crafted on transformers offer the highest transferabilityrate (i.e., 25.7% higher than the average) onto other models. Similarly,adversarial examples crafted on other models have the lowest rate oftransferability (i.e., 56.7% lower than the average) onto transformers. Ourwork emphasizes the importance of studying transformer architectures forattacking and defending models in security domains, and suggests using them asthe primary architecture in transfer attack settings.

The Efficacy of Transformer-based Adversarial Attacks in Security Domains