Investigating the Corruption Robustness of Image Classifiers with Random p-norm Corruptions

Robustness is a fundamental property of machine learning classifiers required to achieve safety and reliability. In the field of adversarial robustness of image classifiers, robustness is commonly defined as the stability of a model to all input changes within a p-norm distance. However, in the field of random corruption robustness, variations observed in the real world are used, while p-norm corruptions are rarely considered. This study investigates the use of random p-norm corruptions to augment the training and test data of image classifiers. We evaluate the model robustness against imperceptible random p-norm corruptions and propose a novel robustness metric. We empirically investigate whether robustness transfers across different p-norms and derive conclusions on which p-norm corruptions a model should be trained and evaluated. We find that training data augmentation with a combination of p-norm corruptions significantly improves corruption robustness, even on top of state-of-the-art data augmentation schemes.

The complete article is a chapter of the book "Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024) - Volume 2: VISAPP".

Bibliographic information

Title:  Investigating the Corruption Robustness of Image Classifiers with Random p-norm Corruptions. 

Written by:  G. Siedel, W. Shao, S. Vock, A. Morozov

in: Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024) - Volume 2: VISAPP Scitepress, 2024.  pages: 171-181, Project number: F 2497, PDF file, DOI: 10.5220/0012397100003660

Download file "Investigating the Corruption Robustness of Image Classifiers with Random p-norm Corruptions" (PDF, 663 KB, Not barrier-free file)

Further Information

Research Project

Project numberF 2497 StatusCompleted Project Safety-related risk assessment of a cyber-physical model system for industry 4.0 applications

To the Project

Research completed