Convolutional Kernel Networks

Julien Mairal 1, * Piotr Koniusz 1, * Zaid Harchaoui 1, * Cordelia Schmid 1, *
* Auteur correspondant
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : An important goal in visual recognition is to devise image representations that are invariant to particular transformations. In this paper, we address this goal with a new type of convolutional neural network (CNN) whose invariance is encoded by a reproducing kernel. Unlike traditional approaches where neural networks are learned either to represent data or for solving a classification task, our network learns to approximate the kernel feature map on training data. Such an approach enjoys several benefits over classical ones. First, by teaching CNNs to be invariant, we obtain simple network architectures that achieve a similar accuracy to more complex ones, while being easy to train and robust to overfitting. Second, we bridge a gap between the neural network literature and kernels, which are natural tools to model invariance. We evaluate our methodology on visual recognition tasks where CNNs have proven to perform well, e.g., digit recognition with the MNIST dataset, and the more challenging CIFAR-10 and STL-10 datasets, where our accuracy is competitive with the state of the art.
Type de document :
Pré-publication, Document de travail
2014
Liste complète des métadonnées

https://hal.inria.fr/hal-01005489
Contributeur : Julien Mairal <>
Soumis le : jeudi 12 juin 2014 - 16:48:59
Dernière modification le : lundi 9 avril 2018 - 12:22:26
Document(s) archivé(s) le : vendredi 12 septembre 2014 - 11:41:05

Fichiers

ckn_arxiv.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01005489, version 1
  • ARXIV : 1406.3332

Citation

Julien Mairal, Piotr Koniusz, Zaid Harchaoui, Cordelia Schmid. Convolutional Kernel Networks. 2014. 〈hal-01005489v1〉

Partager

Métriques

Consultations de la notice

394

Téléchargements de fichiers

771