Indexed Operations for Non-rectangular Lattices Applied to Convolutional Neural Networks - Archive ouverte HAL Access content directly
Conference Papers Year : 2019

Indexed Operations for Non-rectangular Lattices Applied to Convolutional Neural Networks

Abstract

The present paper introduces convolution and pooling operators for indexed images. These operators can be used on images that do not provide Cartesian grids of pixels, as long as a list of neighbor's indices can be provided for each pixel. They are foreseen being useful for convolutional neural networks (CNN) applied to special sensors, especially in science, without requiring image pre-processing. The present work explains the method and its implementation in the Pytorch framework and shows an application of the indexed kernels to the classification task of images with hexagonal lattices using CNN. The obtained results show that the method gives the same performances as the standard convolution kernels. Indexed convolution thus makes deep neural network frameworks more general and capable of addressing unconventional image lattices. The current implementation, as well as code to reproduce the experiments described in this paper are made available as open-source resources on the repository www.github.com/IndexedConv.
Fichier principal
Vignette du fichier
VISAPP_2019_92_CR.pdf (1.17 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02087157 , version 1 (01-04-2019)

Identifiers

  • HAL Id : hal-02087157 , version 1

Cite

Mikaël Jacquemont, Luca Antiga, Thomas Vuillaume, Giorgia Silvestri, A Benoit, et al.. Indexed Operations for Non-rectangular Lattices Applied to Convolutional Neural Networks. VISAPP, 14th International Conference on Computer Vision Theory and Applications, Feb 2019, Prague, Czech Republic. ⟨hal-02087157⟩
341 View
530 Download

Share

Gmail Facebook X LinkedIn More