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

Benefiting from the rapid development of deep learning, 2D and 3D computervision applications are deployed in many safe-critical systems, such asautopilot and identity authentication. However, deep learning models are nottrustworthy enough because of their limited robustness against adversarialattacks. The physically realizable adversarial attacks further pose fatalthreats to the application and human safety. Lots of papers have emerged toinvestigate the robustness and safety of deep learning models againstadversarial attacks. To lead to trustworthy AI, we first construct a generalthreat model from different perspectives and then comprehensively review thelatest progress of both 2D and 3D adversarial attacks. We extend the concept ofadversarial examples beyond imperceptive perturbations and collate over 170papers to give an overview of deep learning model robustness against variousadversarial attacks. To the best of our knowledge, we are the first tosystematically investigate adversarial attacks for 3D models, a flourishingfield applied to many real-world applications. In addition, we examine physicaladversarial attacks that lead to safety violations. Last but not least, wesummarize present popular topics, give insights on challenges, and shed lighton future research on trustworthy AI.

A Survey of Robustness and Safety of 2D and 3D Deep Learning Models Against Adversarial Attacks