Relationship between Uncertainty in DNNs and Adversarial Attacks

TL;DR

This paper reviews how adversarial attacks on deep neural networks can increase their uncertainty, impacting the reliability of DNN predictions across various applications.

Contribution

It provides a comprehensive analysis of the link between DNN uncertainty and adversarial attacks, highlighting how attacks influence model confidence and robustness.

Findings

Adversarial attacks can significantly increase DNN uncertainty.

Uncertainty in DNNs is linked to model and data constraints.

Understanding this relationship can improve robustness strategies.

Abstract

Deep Neural Networks (DNNs) have achieved state of the art results and even outperformed human accuracy in many challenging tasks, leading to DNNs adoption in a variety of fields including natural language processing, pattern recognition, prediction, and control optimization. However, DNNs are accompanied by uncertainty about their results, causing them to predict an outcome that is either incorrect or outside of a certain level of confidence. These uncertainties stem from model or data constraints, which could be exacerbated by adversarial attacks. Adversarial attacks aim to provide perturbed input to DNNs, causing the DNN to make incorrect predictions or increase model uncertainty. In this review, we explore the relationship between DNN uncertainty and adversarial attacks, emphasizing how adversarial attacks might raise DNN uncertainty.