AI Paper Atlas

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

Generalizable Black-Box Adversarial Attack With Meta Learning

Fei Yin, Yong Zhang, Baoyuan Wu, Yan Feng, Jingyi Zhang, Yanbo Fan, Yujiu Yang

TPAMI 2023 · Adversarial robustness and security › Black-box and transfer attacks · approach: Query-efficient black-box attacks · 61 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
add MCG initializationSquare · CIFAR-10 untargetedmean query number: 227.3 -> 39.7Improved
add MCG initializationSquare · CIFAR-10 targetedmean query number: 504.7 -> 92.3Improved
add MCG initializationCG-Attack · ImageNet untargeted WRN-50ASR: 89.6% -> 99.2%Improved
add MCG initializationSimBA-DCT · ImageNet targeted ResNet-18ASR: 56.0% -> 66.8%Improved
add MCG initializationSignhunter · ImageNet targeted Inception-V3ASR: 100.0% -> 99.7%No clear effect
add MCG initializationCG-Attack · JPEG-Compress-WRN-50 defendedASR: 39.0% -> 91.7%Improved
add MCG initializationNES · JPEG-Compress-WRN-50 defendedASR: 13.2% -> 98.9%Improved
add MCG initializationSignhunter · FastAdv-ResNet-50 adversarially trainedASR: 89.4% -> 89.4%No clear effect
add MCG initializationMetaAttack · Imagga tagging APIASR: 16.0% -> 61.0%Improved
add MCG initialization (surrogate and target on different datasets)NES · CIFAR-10 to CIFAR-100 open-setmean query number: 2054.0 -> 773.1Improved
add meta-training in place of plain flow training (ablation)MCG w/ Square · ImageNet ResNet-18FASR: 35.2% -> 57.9%Improved
add surrogate fine-tuning with target query feedback (ablation)MCG w/ Square · ImageNet ResNet-18FASR: 57.9% -> 60.1%Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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