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

Deep neural networks for computer vision are deployed in increasinglysafety-critical and socially-impactful applications, motivating the need toclose the gap in model performance under varied, naturally occurring imagingconditions. Robustness, ambiguously used in multiple contexts includingadversarial machine learning, refers here to preserving model performance undernaturally-induced image corruptions or alterations.  We perform a systematic review to identify, analyze, and summarize currentdefinitions and progress towards non-adversarial robustness in deep learningfor computer vision. We find this area of research has receiveddisproportionately less attention relative to adversarial machine learning, yeta significant robustness gap exists that manifests in performance degradationsimilar in magnitude to adversarial conditions.  Toward developing a more transparent definition of robustness, we provide aconceptual framework based on a structural causal model of the data generatingprocess and interpret non-adversarial robustness as pertaining to a model'sbehavior on corrupted images corresponding to low-probability samples from theunaltered data distribution. We identify key architecture-, data augmentation-,and optimization tactics for improving neural network robustness. Thisrobustness perspective reveals that common practices in the literaturecorrespond to causal concepts. We offer perspectives on how future research maymind this evident and significant non-adversarial robustness gap.

A Systematic Review of Robustness in Deep Learning for Computer Vision: Mind the gap?