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

This paper demonstrates that Non-Maximum Suppression (NMS), which is commonlyused in Object Detection (OD) tasks to filter redundant detection results, isno longer secure. Considering that NMS has been an integral part of OD systems,thwarting the functionality of NMS can result in unexpected or even lethalconsequences for such systems. In this paper, an adversarial example attackwhich triggers malfunctioning of NMS in end-to-end OD models is proposed. Theattack, namely \texttt{Daedalus}, compresses the dimensions of detection boxesto evade NMS. As a result, the final detection output contains extremely densefalse positives. This can be fatal for many OD applications such as autonomousvehicles and surveillance systems. The attack can be generalised to differentend-to-end OD models, such that the attack cripples various OD applications.Furthermore, a way to craft robust adversarial examples is developed by usingan ensemble of popular detection models as the substitutes. Considering thepervasive nature of model reusing in real-world OD scenarios, Daedalus examplescrafted based on an \textit{ensemble of substitutes} can launch attacks withoutknowing the parameters of the victim models. Experimental results demonstratethat the attack effectively stops NMS from filtering redundant bounding boxes.As the evaluation results suggest, Daedalus increases the false positive ratein detection results to $99.9\%$ and reduces the mean average precision scoresto $0$, while maintaining a low cost of distortion on the original inputs. Itis also demonstrated that the attack can be practically launched againstreal-world OD systems via printed posters.

Daedalus: Breaking Non-Maximum Suppression in Object Detection via Adversarial Examples