Learning Graphs to Match

Minsu Cho 1, 2 Karteek Alahari 2, 1, 3 Jean Ponce 2, 1
2 WILLOW - Models of visual object recognition and scene understanding
DI-ENS - Département d'informatique de l'École normale supérieure, Inria Paris-Rocquencourt, CNRS - Centre National de la Recherche Scientifique : UMR8548
3 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : Many tasks in computer vision are formulated as graph matching problems. Despite the NP-hard nature of the problem, fast and accurate approximations have led to significant progress in a wide range of applications. Learning graph models from observed data, however, still remains a challenging issue. This paper presents an effective scheme to parameterize a graph model, and learn its structural attributes for visual object matching. For this, we propose a graph representation with histogram-based attributes, and optimize them to increase the matching accuracy. Experimental evaluations on synthetic and real image datasets demonstrate the effectiveness of our approach, and show significant improvement in matching accuracy over graphs with pre-defined structures.
Type de document :
Communication dans un congrès
ICCV - IEEE International Conference on Computer Vision, Dec 2013, Sydney, Australia. IEEE, pp.25-32, 2013, 〈10.1109/ICCV.2013.11〉
Liste complète des métadonnées

Littérature citée [31 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-00875105
Contributeur : Karteek Alahari <>
Soumis le : lundi 21 octobre 2013 - 10:19:46
Dernière modification le : lundi 28 mai 2018 - 15:10:02
Document(s) archivé(s) le : vendredi 7 avril 2017 - 13:29:48

Fichier

cho13.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Minsu Cho, Karteek Alahari, Jean Ponce. Learning Graphs to Match. ICCV - IEEE International Conference on Computer Vision, Dec 2013, Sydney, Australia. IEEE, pp.25-32, 2013, 〈10.1109/ICCV.2013.11〉. 〈hal-00875105〉

Partager

Métriques

Consultations de la notice

890

Téléchargements de fichiers

405