Boosting Adversarial Attacks on Neural Networks with Better Optimizer

TL;DR

This paper introduces a modified Adam optimizer combined with iterative gradient attacks to significantly enhance the transferability and success rate of adversarial examples against neural networks, especially in black-box settings.

Contribution

It proposes the Adam Iterative Fast Gradient Method, a novel approach that improves adversarial attack success rates and transferability over existing methods.

Findings

Achieved a 95.0% attack success rate on defense models.

Outperformed existing iterative attack methods in experiments.

Enhanced transferability of adversarial examples in black-box environments.

Abstract

Convolutional neural networks have outperformed humans in image recognition tasks, but they remain vulnerable to attacks from adversarial examples. Since these data are crafted by adding imperceptible noise to normal images, their existence poses potential security threats to deep learning systems. Sophisticated adversarial examples with strong attack performance can also be used as a tool to evaluate the robustness of a model. However, the success rate of adversarial attacks can be further improved in black-box environments. Therefore, this study combines a modified Adam gradient descent algorithm with the iterative gradient-based attack method. The proposed Adam Iterative Fast Gradient Method is then used to improve the transferability of adversarial examples. Extensive experiments on ImageNet showed that the proposed method offers a higher attack success rate than existing iterative methods. By extending our method, we achieved a state-of-the-art attack success rate of 95.0% on defense models.