Stochastic sparse adversarial attacks
Not on the map. Why?
Abstract (arXiv)
This paper introduces stochastic sparse adversarial attacks (SSAA), standing as simple, fast and purely noise-based targeted and untargeted attacks of neural network classifiers (NNC). SSAA offer new examples of sparse (or L0) attacks for which only few methods have been proposed previously. These attacks are devised by exploiting a small-time expansion idea widely used for Markov processes. Experiments on small and large datasets (CIFAR-10 and ImageNet) illustrate several advantages of SSAA in comparison with the-state-of-the-art methods. For instance, in the untargeted case, our method called Voting Folded Gaussian Attack (VFGA) scales efficiently to ImageNet and achieves a significantly lower L0 score than SparseFool (up to frac25) while being faster. Moreover, VFGA achieves better L0 scores on ImageNet than Sparse-RS when both attacks are fully successful on a large number of samples.
Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). First posted on arXiv 24 Nov 2020.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2021: 1; 2022: 1; 2023: 0; 2024: 2.
4 papers.
- 2024AttackBench: Evaluating Gradient-based Attacks for Adversarial Examplescited by 14
- 2024σ-zero: Gradient-based Optimization of ℓ0-norm Adversarial Examplescited by 7
- 2021Improving Robustness of Deep Reinforcement Learning Agents: Environment Attack based on the Critic Networkcited by 3
- 2022Riemannian data-dependent randomized smoothing for neural networks certificationcited by 1
- 2024AttackBench: Evaluating Gradient-based Attacks for Adversarial Examples30 Apr 2024
- 2024σ-zero: Gradient-based Optimization of ℓ0-norm Adversarial Examples2 Feb 2024
- 2022Riemannian data-dependent randomized smoothing for neural networks certification21 Jun 2022
- 2021Improving Robustness of Deep Reinforcement Learning Agents: Environment Attack based on the Critic Network7 Apr 2021