AI Paper Atlas

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

S 4 ST: A Strong, Self-Transferable, faSt, and Simple Scale Transformation for Data-Free Transferable Targeted Attack

Yongxiang Liu, Bowen Peng, Li Liu, Xiang Li

TPAMI 2026 · Adversarial robustness and security › Black-box and transfer attacks · approach: Data-free substitute training · 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
add S4ST-Aug (complementary augmentation pool) (ablation)S4ST-Base · ResNet-50 surrogateaverage tSuc: 58.2 -> 76.8Improved
add S4ST-Block (block-wise scaling, m=6) (ablation)S4ST-Base · ResNet-50 surrogateaverage tSuc: 58.2 -> 74.3Improved
combine Aug and Block (ablation)S4ST-Base · ResNet-50 surrogateaverage tSuc: 58.2 -> 83.0Improved
add flip to augmentation pool (ablation)S4ST · ResNet-50 surrogateaverage tSuc: 80.8 -> 83.0Improved
add contrast to augmentation pool (ablation)S4ST · ResNet-50 surrogateaverage tSuc: 82.6 -> 83.0No clear effect
add brightness to augmentation pool (ablation)S4ST · ResNet-50 surrogateaverage tSuc: 82.7 -> 83.0No clear effect
add saturation to augmentation pool (ablation)S4ST · ResNet-50 surrogateaverage tSuc: 82.7 -> 83.0No clear effect
add hue to augmentation pool (ablation)S4ST · ResNet-50 surrogateaverage tSuc: 82.6 -> 83.0No clear effect
use scaling instead of shear as the basic transformationTMI · ResNet-50 surrogateaverage tSuc: 48.8% -> 58.0%Improved
tune DI and RDI parametersDI / RDI · ResNet-50 surrogateaverage tSuc: 22.4% -> 36.6% (DI), 27.3% -> 41.7% (RDI)Improved
use bidirectional (upscale and downscale) scalingDI/RDI · ResNet-50 surrogateaverage tSuc: 41.7% -> 56.8%Improved
replace simple bidirectional scaling with S4ST-Basescaling · ResNet-50 surrogateaverage tSuc: 57.2% -> 61.6%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