Universal adversarial perturbations for multiple classification tasks with quantum classifiers

TL;DR

This paper investigates quantum universal adversarial perturbations, demonstrating that a single perturbation can deceive quantum classifiers across different tasks, highlighting vulnerabilities in quantum machine learning models.

Contribution

It introduces the concept of quantum universal adversarial perturbations for heterogeneous classification tasks and shows their effectiveness on real-world datasets.

Findings

A single universal perturbation can deceive classifiers on multiple tasks.

Quantum classifiers with high accuracy are vulnerable to crafted perturbations.

The method applies to datasets like handwritten digits and MRI images.

Abstract

Quantum adversarial machine learning is an emerging field that studies the vulnerability of quantum learning systems against adversarial perturbations and develops possible defense strategies. Quantum universal adversarial perturbations are small perturbations, which can make different input samples into adversarial examples that may deceive a given quantum classifier. This is a field that was rarely looked into but worthwhile investigating because universal perturbations might simplify malicious attacks to a large extent, causing unexpected devastation to quantum machine learning models. In this paper, we take a step forward and explore the quantum universal perturbations in the context of heterogeneous classification tasks. In particular, we find that quantum classifiers that achieve almost state-of-the-art accuracy on two different classification tasks can be both conclusively deceived by one carefully-crafted universal perturbation. This result is explicitly demonstrated with well-designed quantum continual learning models with elastic weight consolidation method to avoid catastrophic forgetting, as well as real-life heterogeneous datasets from hand-written digits and medical MRI images. Our results provide a simple and efficient way to generate universal perturbations on heterogeneous classification tasks and thus would provide valuable guidance for future quantum learning technologies.