Chromatic and spatial analysis of one-pixel attacks against an image classifier

TL;DR

This paper investigates the mechanisms of one-pixel attacks on neural network classifiers, analyzing chromatic and spatial patterns, and introduces confidence maps to better understand attack success and classifier behavior.

Contribution

It provides a detailed analysis of one-pixel attack characteristics, including chromatic and spatial distributions, and introduces confidence maps to visualize classifier responses.

Findings

Successful attacks tend to alter pixel color more.

Effective attacks are often located at the image center.

Analysis helps understand classifier vulnerabilities.

Abstract

One-pixel attack is a curious way of deceiving neural network classifier by changing only one pixel in the input image. The full potential and boundaries of this attack method are not yet fully understood. In this research, the successful and unsuccessful attacks are studied in more detail to illustrate the working mechanisms of a one-pixel attack created using differential evolution. The data comes from our earlier studies where we applied the attack against medical imaging. We used a real breast cancer tissue dataset and a real classifier as the attack target. This research presents ways to analyze chromatic and spatial distributions of one-pixel attacks. In addition, we present one-pixel attack confidence maps to illustrate the behavior of the target classifier. We show that the more effective attacks change the color of the pixel more, and that the successful attacks are situated at the center of the images. This kind of analysis is not only useful for understanding the behavior of the attack but also the qualities of the classifying neural network.