Designing Affine-Invariant Neural Networks for Photometric Corruption Robustness and Generalization

Mounir Messaoudi, Quentin Rapilly, Sébastien Herbreteau, Anaïs Badoual, Charles Kervrann

Keywords: Computer Vision, Robust neural network, invariance, equivariance, biological imaging, microscopy, classification, object localization

TL;DR: Scale-Equivariant & Shift-Invariant NN rendered Affine invariant for application to image classification and object localization

Abstract: Standard Convolutional Neural Networks are notoriously sensitive to photometric variations, a critical flaw that data augmentation only partially mitigates without offering formal guarantees. We introduce the *Scale-Equivariant Shift-Invariant* (*SEqSI*) model, a novel architecture that achieves intensity scale equivariance and intensity shift invariance by design, enabling full invariance to global intensity affine transformations with appropriate post-processing. By strategically prepending a single shift-invariant layer to a scale-equivariant backbone, *SEqSI* provides these formal guarantees while remaining fully compatible with common components like ReLU. We benchmark *SEqSI* against *Standard*, *Scale-Equivariant* (*SEq*), and *Affine-Equivariant* (*AffEq*) models on 2D and 3D image-classification and object-localization tasks. Our experiments demonstrate that *SEqSI* architectural properties provide certified robustness to affine intensity transformations and enhances generalization across non-affine corruptions and domain shifts in challenging real-world applications like biological image analysis. This work establishes *SEqSI* as a practical and principled approach for building photometrically robust models without major trade-offs.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 12601