On learning to localize objects with minimal supervision

Hyun Oh Song 1 Ross Girshick 1 Stefanie Jegelka 2 Julien Mairal 3, * Zaid Harchaoui 3 Trevor Darrell 1
* Auteur correspondant
3 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : Learning to localize objects with minimal supervision is an important problem in computer vision, since large fully annotated datasets are extremely costly to obtain. In this paper, we propose a new method that achieves this goal with only image-level labels of whether the objects are present or not. Our approach combines a discriminative submodular cover problem for automatically discovering a set of positive object windows with a smoothed latent SVM formulation. The latter allows us to leverage efficient quasi Newton optimization techniques. Our experiments demonstrate that the proposed approach provides a 50% relative improvement in mean average precision over the current state-of-the-art on PASCAL VOC 2007 detection.
Type de document :
Communication dans un congrès
Tony Jebara; Eric P. Xing. ICML - 31st International Conference on Machine Learning, Jun 2014, Beijing, China. JMLR, 32, pp.1611-1619, 2014, JMLR Workshop and Conference Proceedings
Liste complète des métadonnées


https://hal.inria.fr/hal-00996849
Contributeur : Julien Mairal <>
Soumis le : mardi 27 mai 2014 - 09:38:37
Dernière modification le : samedi 18 février 2017 - 01:07:06
Document(s) archivé(s) le : mercredi 27 août 2014 - 11:01:21

Fichier

hyun_icml14.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00996849, version 1

Collections

Citation

Hyun Oh Song, Ross Girshick, Stefanie Jegelka, Julien Mairal, Zaid Harchaoui, et al.. On learning to localize objects with minimal supervision. Tony Jebara; Eric P. Xing. ICML - 31st International Conference on Machine Learning, Jun 2014, Beijing, China. JMLR, 32, pp.1611-1619, 2014, JMLR Workshop and Conference Proceedings. <hal-00996849>

Partager

Métriques

Consultations de
la notice

1302

Téléchargements du document

645