Keuper Labs
← All publications

Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops (ICCVw 2023) · 2023

On the Interplay of Convolutional Padding and Adversarial Robustness

Paul Gavrikov, Janis Keuper

Adversarial RobustnessCNNsSignal Processing

Why this publication matters

The extra pixels added around an image before processing may seem like a minor implementation choice. This paper shows that they can influence where attacks concentrate and how well a network resists them. It brings a usually hidden design decision into the discussion of dependable image recognition.

Abstract

It is common practice to apply padding prior to convolution operations to preserve the resolution of feature-maps in Convolutional Neural Networks (CNN). While many alternatives exist, this is often achieved by adding a border of zeros around the inputs. In this work, we show that adversarial attacks often result in perturbation anomalies at the image boundaries, which are the areas where padding is used. Consequently, we aim to provide an analysis of the interplay between padding and adversarial attacks and seek an answer to the question of how different padding modes (or their absence) affect adversarial robustness in various scenarios.

Abstract source ↗

Figures

Spatial distribution of adversarial perturbations for different padding methods.
Figure 1. Average perturbations of 1,000 CIFAR-10 samples where attacks were successful. Adversarial attacks show perturbation anomalies (more or fewer perturbations) at image boundaries that can be attributed to padding. Shown here is a ResNet-20 trained without adversarial defenses with different padding modes under attacks of highbudget l∞- (top row) and l2-bounded (bottom row) APGD- CE. View in source ↗
Examples of a CIFAR-10 training sample increased to an image resolution to 64 × 64 px by different padding approaches.
Figure 2. Examples of a CIFAR-10 training sample increased to an image resolution to 64 × 64 px by different padding approaches. View in source ↗

Cite this paper

Download .bib 
@inproceedings{gavrikov2023ontheinterplay32,
  title = {{On the Interplay of Convolutional Padding and Adversarial Robustness}},
  author = {Paul Gavrikov and Janis Keuper},
  booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops (ICCVw 2023)},
  year = {2023},
  url = {https://arxiv.org/pdf/2308.06612}
}

Figures and abstract are reproduced from the linked research sources. Credit remains with the authors and publishers.