Human Imperceptible Attacks and Applications to Improve Fairness

TL;DR

This paper introduces a DRO framework that creates human-imperceptible attacks on neural networks, which not only degrade performance more effectively but also help improve fairness in image classification tasks.

Contribution

The paper presents a novel DRO-based attack method that generates less perceptible adversarial examples and uses these to enhance fairness in neural network classification.

Findings

Our attack algorithm outperforms existing methods in imperceptibility.

DRO training with these attacks improves fairness in image classification.

The proposed algorithmic implementation accelerates DRO training significantly.

Abstract

Modern neural networks are able to perform at least as well as humans in numerous tasks involving object classification and image generation. However, small perturbations which are imperceptible to humans may significantly degrade the performance of well-trained deep neural networks. We provide a Distributionally Robust Optimization (DRO) framework which integrates human-based image quality assessment methods to design optimal attacks that are imperceptible to humans but significantly damaging to deep neural networks. Through extensive experiments, we show that our attack algorithm generates better-quality (less perceptible to humans) attacks than other state-of-the-art human imperceptible attack methods. Moreover, we demonstrate that DRO training using our optimally designed human imperceptible attacks can improve group fairness in image classification. Towards the end, we provide an algorithmic implementation to speed up DRO training significantly, which could be of independent interest.