Reinforcement Learning Platform for Adversarial Black-box Attacks with Custom Distortion Filters

TL;DR

This paper introduces RLAB, a reinforcement learning platform for generating adversarial examples with customizable distortions, improving attack efficiency and robustness evaluation of image classifiers.

Contribution

The paper presents a novel RL-based platform with a dual-action method for efficient black-box adversarial attacks and robustness assessment.

Findings

RLAB outperforms state-of-the-art methods in query efficiency.

Retraining with adversarial samples enhances model robustness.

The platform enables targeted and untargeted attacks with various distortion filters.

Abstract

We present a Reinforcement Learning Platform for Adversarial Black-box untargeted and targeted attacks, RLAB, that allows users to select from various distortion filters to create adversarial examples. The platform uses a Reinforcement Learning agent to add minimum distortion to input images while still causing misclassification by the target model. The agent uses a novel dual-action method to explore the input image at each step to identify sensitive regions for adding distortions while removing noises that have less impact on the target model. This dual action leads to faster and more efficient convergence of the attack. The platform can also be used to measure the robustness of image classification models against specific distortion types. Also, retraining the model with adversarial samples significantly improved robustness when evaluated on benchmark datasets. The proposed platform outperforms state-of-the-art methods in terms of the average number of queries required to cause misclassification. This advances trustworthiness with a positive social impact.