AI Paper Atlas

Research map / Adversarial robustness and security / Black-box and transfer attacks

Blurred-Dilated Method for Adversarial Attacks

Yang Deng, Weibin Wu, Jianping Zhang, Zibin Zheng

NeurIPS 2023 · Adversarial robustness and security › Black-box and transfer attacks · approach: Transferable adversarial examples · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace RN50 source with BD RN50source model for MI-FGSM transfer attacks · ImageNet, eps=16/255average attack success rate: 72.8% -> 95.9%Improved
replace VGG16 source with BD VGG16source model for MI-FGSM transfer attacks · ImageNet, eps=16/255average attack success rate: 72.8% -> 86.4%Improved
replace DN121 source with BD DN121source model for MI-FGSM transfer attacks · ImageNet, eps=16/255average attack success rate: 76.2% -> 88.3%Improved
replace MobV2 source with BD MobV2source model for MI-FGSM transfer attacks · ImageNet, eps=16/255average attack success rate: 61.2% -> 82.3%Improved
BD vs LinBP, the strongest baseline, with RN50 sourcesource model for MI-FGSM transfer attacks · ImageNet, eps=16/255average black-box success rate: 84.2% -> 95.3%Improved
BD vs LinBP at small budgetsource model for MI-FGSM transfer attacks · ImageNet, eps=4/255average black-box success rate: 35.8% -> 56.8%Improved
add BD to ILAsource model for MI-FGSM transfer attacks · ImageNet, eps=16/255average black-box success rate: 78.4% -> 96.3%Improved
add BD to LinBPsource model for MI-FGSM transfer attacks · ImageNet, eps=16/255average black-box success rate: 84.2% -> 97.3%Improved
add BD to ILA+LinBPsource model for MI-FGSM transfer attacks · ImageNet, eps=16/255average black-box success rate: 88.7% -> 97.9%Improved
remove dilated convolutions (BlurPool only) (ablation)source model for MI-FGSM transfer attacks · ImageNet, eps=16/255average black-box success rate: 91.0% -> 85.4%Hurt
remove BlurPool (dilated conv only) (ablation)source model for MI-FGSM transfer attacks · ImageNet, eps=16/255average black-box success rate: 91.0% -> 69.9%Hurt
BD ResNet-20 source on low-resolution datasource model for MI-FGSM transfer attacks · CIFAR-10, eps=16/255average attack success rate: 95.2% -> 97.2%Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy