Generalizing universal adversarial perturbations for deep neural networks

DOI10.1007/S10994-023-06306-ZMaRDI QIDQ6103573zbMATH OpenFDO

Authors Yanghao Zhang, Wenjie Ruan, Fu Wang, Xiaowei Huang

Publication date 27 June 2023

Published in Machine Learning (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/S10994-023-06306-Z

zbMATH Keywords

adversarial examplesdeep learningdeep neural networkssecurity

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Recommendations

• Generating universal adversarial perturbation with ResNet
• Enhancing adversarial attack transferability with multi-scale feature attack
• Black-box adversarial attacks by manipulating image attributes
• Adversarial perturbations of deep neural networks
• LADDER: latent boundary-guided adversarial training

Cites work

• A game-based approximate verification of deep neural networks with provable guarantees
• A survey of safety and trustworthiness of deep neural networks: verification, testing, adversarial attack and defence, and interpretability
• Handbook on Computer Learning and Intelligence
• Natural language processing (almost) from scratch
• Nonlinear total variation based noise removal algorithms

Cited in

(6)

• D3AdvM: A direct 3D adversarial sample attack inside mesh data
• Adversarial Framework with Certified Robustness for Time-Series Domain via Statistical Features
• Metrics and methods for robustness evaluation of neural networks with generative models
• Salient feature extractor for adversarial defense on deep neural networks
• Mappings, dimensionality and reversing out of deep neural networks
• The feasibility and inevitability of stealth attacks

This page was built for publication: Generalizing universal adversarial perturbations for deep neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6103573)