Metric Learning for Large Scale Image Classification: Generalizing to New Classes at Near-Zero Cost

Thomas Mensink 1, 2 Jakob Verbeek 1 Florent Perronnin 2 Gabriela Csurka 2
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : We are interested in large-scale image classification and especially in the setting where images corresponding to new or existing classes are continuously added to the training set. Our goal is to devise classifiers which can incorporate such images and classes on-the-fly at (near) zero cost. We cast this problem into one of learning a metric which is shared across all classes and explore k-nearest neighbor (k-NN) and nearest class mean (NCM) classifiers. We learn metrics on the ImageNet 2010 challenge data set, which contains more than 1.2M training images of 1K classes. Surprisingly, the NCM classifier compares favorably to the more flexible k-NN classifier, and has comparable performance to linear SVMs. We also study the generalization performance, among others by using the learned metric on the ImageNet-10K dataset, and we obtain competitive performance. Finally, we explore zero-shot classification, and show how the zero-shot model can be combined very effectively with small training datasets.
Type de document :
Communication dans un congrès
Andrew Fitzgibbon and Svetlana Lazebnik and Pietro Perona and Yoichi Sato and Cordelia Schmid. ECCV 2012 - 12th European Conference on Computer Vision, Oct 2012, Florence, Italy. Springer, 7573, pp.488-501, 2012, Lecture Notes in Computer Science. 〈10.1007/978-3-642-33709-3_35〉
Liste complète des métadonnées

Littérature citée [28 références]  Voir  Masquer  Télécharger


https://hal.inria.fr/hal-00722313
Contributeur : Thoth Team <>
Soumis le : mercredi 1 août 2012 - 11:41:55
Dernière modification le : mercredi 11 avril 2018 - 01:58:29
Document(s) archivé(s) le : vendredi 2 novembre 2012 - 02:30:50

Fichiers

mensink12eccv.final.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Thomas Mensink, Jakob Verbeek, Florent Perronnin, Gabriela Csurka. Metric Learning for Large Scale Image Classification: Generalizing to New Classes at Near-Zero Cost. Andrew Fitzgibbon and Svetlana Lazebnik and Pietro Perona and Yoichi Sato and Cordelia Schmid. ECCV 2012 - 12th European Conference on Computer Vision, Oct 2012, Florence, Italy. Springer, 7573, pp.488-501, 2012, Lecture Notes in Computer Science. 〈10.1007/978-3-642-33709-3_35〉. 〈hal-00722313〉

Partager

Métriques

Consultations de la notice

1620

Téléchargements de fichiers

8047