Large-scale image classification with trace-norm regularization

Zaid Harchaoui 1 Matthijs Douze 1, 2 Mattis Paulin 1 Miro Dudik 3 Jérôme Malick 4
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
3 NY
Yahoo! Labs New York
4 BIPOP - Modelling, Simulation, Control and Optimization of Non-Smooth Dynamical Systems
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : With the advent of larger image classification datasets such as ImageNet, designing scalable and efficient multi-class classification algorithms is now an important challenge. We introduce a new scalable learning algorithm for large-scale multi-class image classification, based on the multinomial logistic loss and the trace-norm regularization penalty. Reframing the challenging non-smooth optimization problem into a surrogate infinite-dimensional optimization problem with a regular l1 -regularization penalty, we propose a simple and provably efficient accelerated coordinate descent algorithm. Furthermore, we show how to perform efficient matrix computations in the compressed domain for quantized dense visual features, scaling up to 100,000s examples, 1,000s-dimensional features, and 100s of categories. Promising experimental results on the "Fungus", "Ungulate", and "Vehicles" subsets of ImageNet are presented, where we show that our approach performs significantly better than state-of-the-art approaches for Fisher vectors with 16 Gaussians.
Type de document :
Communication dans un congrès
CVPR - IEEE Conference on Computer Vision & Pattern Recognition, Jun 2012, Providence, United States. IEEE, pp.3386-3393, 2012, 〈10.1109/CVPR.2012.6248078〉
Liste complète des métadonnées

Littérature citée [31 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-00728388
Contributeur : Thoth Team <>
Soumis le : mercredi 5 septembre 2012 - 17:56:09
Dernière modification le : mercredi 11 avril 2018 - 01:58:26
Document(s) archivé(s) le : vendredi 16 décembre 2016 - 10:22:31

Fichier

hdpdm_2012_ultrarod.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Citation

Zaid Harchaoui, Matthijs Douze, Mattis Paulin, Miro Dudik, Jérôme Malick. Large-scale image classification with trace-norm regularization. CVPR - IEEE Conference on Computer Vision & Pattern Recognition, Jun 2012, Providence, United States. IEEE, pp.3386-3393, 2012, 〈10.1109/CVPR.2012.6248078〉. 〈hal-00728388〉

Partager

Métriques

Consultations de la notice

1056

Téléchargements de fichiers

808