On the Shift Invariance of Max Pooling Feature Maps in Convolutional Neural Networks - Statistique pour le Vivant et l’Homme
Pré-Publication, Document De Travail Année : 2023

On the Shift Invariance of Max Pooling Feature Maps in Convolutional Neural Networks

Résumé

This paper focuses on improving the mathematical interpretability of convolutional neural networks (CNNs) in the context of image classification. Specifically, we tackle the instability issue arising in their first layer, which tends to learn parameters that closely resemble oriented band-pass filters when trained on datasets like ImageNet. Subsampled convolutions with such Gabor-like filters are prone to aliasing, causing sensitivity to small input shifts. In this context, we establish conditions under which the max pooling operator approximates a complex modulus, which is nearly shift invariant. We then derive a measure of shift invariance for subsampled convolutions followed by max pooling. In particular, we highlight the crucial role played by the filter's frequency and orientation in achieving stability. We experimentally validate our theory by considering a deterministic feature extractor based on the dual-tree complex wavelet packet transform, a particular case of discrete Gabor-like decomposition.
Fichier principal
Vignette du fichier
preprint.pdf (1.71 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03779434 , version 1 (16-09-2022)
hal-03779434 , version 2 (24-10-2023)

Licence

Identifiants

  • HAL Id : hal-03779434 , version 2

Citer

Hubert Leterme, Kévin Polisano, Valérie Perrier, Karteek Alahari. On the Shift Invariance of Max Pooling Feature Maps in Convolutional Neural Networks. 2023. ⟨hal-03779434v2⟩
343 Consultations
106 Téléchargements

Partager

More