Attack-Dependent Composition Effects and Defense Selection in Adversarial Image Classification
Mathematics · 16.11.2026 · 14 (22) · 4213
DOI: 10.3390/math14224213 ↗Kurz gefasst
Fachartikel in der Zeitschrift Mathematics, online erschienen am 16.11.2026. Verfasst von Nataliia Kussul und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Education and Science, Horizon Europe project FUTUREFOR.
Abstract (Originalsprache)
Input-preprocessing defenses are often combined, and studies of such combinations ask whether a composition improves robustness. When the effect of adding a component changes sign across attacks, this leaves open which pipeline should be deployed. We treat the choice among preprocessing pipelines as a decision problem over a finite table of recovery rates and characterize its geometry: a pipeline is optimal for some attack weights exactly when no mixture of the other pipelines strictly dominates it, each pipeline’s winning region is a convex polytope, and preferences switch on hyperplanes. We apply this framework to Gaussian perturbation, feature squeezing, spatial smoothing, and their pairwise compositions, using DenseNet169 and ResNet-50 on ImageNetV2 (614 and 569 correctly classified images from as many classes) and four attacks generated against the undefended classifiers. On both architectures, adding Gaussian noise after median smoothing raises recovery under PGD, FGSM, and BIM by 7.0–16.4 percentage points but lowers it under Boundary Attack by 22.1–27.6 points relative to the better constituent. Mean and maximin recovery select this composition, whereas minimax regret selects smoothing alone. On both architectures, removing dominated pipelines discards the identity and feature squeezing with smoothing; for DenseNet169, feature squeezing with Gaussian perturbation wins no individual attack yet is the unique optimum for mixtures of FGSM and Boundary Attack, and it remains optimal for some attack weights in 96% of bootstrap replicates. When clean accuracy is weighted by an attack probability, no preprocessing is preferred below a probability of about 0.26 for DenseNet169 and 0.42 for ResNet-50. An adaptive BPDA/EOT attack reduces the recovery of every tested pipeline to zero, so the empirical results describe a non-adaptive threat model.
Förderung
- Ministry of Education and Science
- Horizon Europe project FUTUREFOR
Zitieren
APA
Kussul, N., Omelchenko, B., & Shelestov, A. (2026). Attack-Dependent Composition Effects and Defense Selection in Adversarial Image Classification. Mathematics, 14(22), 4213. https://doi.org/10.3390/math14224213
Vancouver
Kussul N, Omelchenko B, Shelestov A. Attack-Dependent Composition Effects and Defense Selection in Adversarial Image Classification. Mathematics. 2026;14(22):4213. doi:10.3390/math14224213
BibTeX
@article{Kussul2026,
author = {Kussul, Nataliia and Omelchenko, Bohdan and Shelestov, Andrii},
title = {Attack-Dependent Composition Effects and Defense Selection in Adversarial Image Classification},
journal = {Mathematics},
year = {2026},
volume = {14},
number = {22},
pages = {4213},
doi = {10.3390/math14224213}
}
EUVIMED