SAIF: Sparse Adversarial and Imperceptible Attack Framework

2022/12/14 by Tooba Imtiaz, Morgan Kohler, Imtiaz, Tooba +12 · 1 citation
Computer Science · Medicine · #Adversarial Robustness in Machine Learning #Artificial Intelligence in Healthcare and Education #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2212.07495

openalex publication_date 2022/12/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Adversarial attacks hamper the decision-making ability of neural networks by perturbing the input signal. The addition of calculated small distortion to images, for instance, can deceive a well-trained image classification network. In this work, we propose a novel attack technique called Sparse Adversarial and Interpretable Attack Framework (SAIF). Specifically, we design imperceptible attacks that contain low-magnitude perturbations at a small number of pixels and leverage these sparse attacks to reveal the vulnerability of classifiers. We use the Frank-Wolfe (conditional gradient) algorithm to simultaneously optimize the attack perturbations for bounded magnitude and sparsity with O(1/√(T)) convergence. Empirical results show that SAIF computes highly imperceptible and interpretable adversarial examples, and outperforms state-of-the-art sparse attack methods on the ImageNet dataset.

Cited by

• ZQBA: Zero Query Black-box Adversarial Attack

Related

• Learning to Attack with Fewer Pixels: A Probabilistic Post-hoc Framework for Refining Arbitrary Dense Adversarial Attacks
• Sparse and Imperceivable Adversarial Attacks
• A Versatile Framework for Designing Group-Sparse Adversarial Attacks
• Efficient and Robust Classification for Sparse Attacks
• GSE: Group-wise Sparse and Explainable Adversarial Attacks