AI Paper Atlas

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

Learning with Bilevel-Minimax Optimization for Efficient and Reliable Transfer Attacks

Yaohua Liu, Yifan Guo, Jiaxin Gao

ECCV 2026 · Adversarial robustness and security › Black-box and transfer attacks · approach: Transferable adversarial examples · 0 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 BMAT to 9 base attackers (24 combinations)PGD/MI/VMI/SI/DI/TI + SGM/Ghost/MBA · ImageNet (ResNet-50 surrogate)average ASR: average ASR gain of 23.28%Improved
use BMAT instead of RAP at equal backward passesRAP · ImageNet BP-normalizedaverage ASR: 5.44 -> 15.03Improved
give RAP 10x more backward passes (BP=400)RAP · ImageNet BP-normalizedaverage ASR: 9.68 vs BMAT 15.03Mixed
use BMAT instead of BETAK initialization learningBETAK · ImageNet BP-normalizedaverage ASR and memory: 12.06 -> 15.03 ASR; 22.69 -> 7.89 GBImproved
add IGA to MI (ablation)MI · Cityscapes mIoU, FCN surrogate, UPerNet victimmIoU: 3.57 -> 2.40Improved
add SWM on top of IGA (ablation)MI + IGA · Cityscapes mIoU, DLV3-R50 surrogate, Setr victimmIoU: 44.42 -> 42.61Improved
add SWM on top of IGA (ablation)MI + IGA · Cityscapes mIoU, DLV3-R50 surrogate, UPerNet victimmIoU: 1.56 -> 1.96Hurt
increase PGD iterations from K=10 to K=40 (ablation)PGD · ImageNet, IncRes-V2 victimASR: 13.64 -> 13.22Hurt
add BMAT with (T,K~,K)=(3,10,10) against PGD with more iterations (ablation)PGD · ImageNet attack iterationsaverage ASR gain: +25.87% at K=20, +41.45% at K=40Improved
keep SWM in Phase-II (Full BMAT instead of Fast) (ablation)BMAT Single-Surrogate on DI-MI · ImageNetaverage ASR: 46.53 -> 49.25Improved
use auxiliary Inc-v3 pseudo-surrogate instead of Bayesian single surrogate (ablation)BMAT Fast on DI-MI · ImageNetaverage ASR: 46.53 -> 49.37Improved
use single-surrogate Bayesian pseudo-surrogate (no extra victim access) (ablation)BMAT · ImageNetaverage ASR: 30.17% and 58.34% average ASR improvement over baselinesImproved

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