Black-box adversarial attacks by manipulating image attributes

DOI10.1016/J.INS.2020.10.028MaRDI QIDQ2123528zbMATH OpenOpenAlexFDO

Authors X. X. Wei, Ying Guo, Bo Li

Publication date 14 April 2022

Published in Information Sciences (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/j.ins.2020.10.028

zbMATH Keywords

adversarial attackadversarial attributesblack-box setting

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Cryptography (94A60)

Recommendations

• Enhancing adversarial attack transferability with multi-scale feature attack
• An empirical study of derivative-free-optimization algorithms for targeted black-box attacks in deep neural networks
• Understanding adversarial robustness via critical attacking route
• Spanning attack: reinforce black-box attacks with unlabeled data
• Adversarial defense via the data-dependent activation, total variation minimization, and adversarial training

Cites work

• Analysis of classifiers' robustness to adversarial perturbations
• Differential evolution -- a simple and efficient heuristic for global optimization over continuous spaces

Cited in

(11)

• Understanding adversarial robustness via critical attacking route
• An empirical study of derivative-free-optimization algorithms for targeted black-box attacks in deep neural networks
• Spanning attack: reinforce black-box attacks with unlabeled data
• Generating universal adversarial perturbation with ResNet
• Greedy attack and Gumbel attack: generating adversarial examples for discrete data
• Enhancing adversarial attack transferability with multi-scale feature attack
• Detecting scene-plausible perceptible backdoors in trained DNNs without access to the training set
• Attacking the tracker with a universal and attractive patch as fake target
• Generalizing universal adversarial perturbations for deep neural networks
• Salient feature extractor for adversarial defense on deep neural networks
• An adversarial attack based on multi-objective optimization in the black-box scenario: MOEA-APGA II

Describes a project that uses

Uses Software

• SemanticAdv
• DeepFool
• ImageNet
• AlexNet
• CIFAR
• Grad-CAM

This page was built for publication: Black-box adversarial attacks by manipulating image attributes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2123528)