Seeing 3D chairs: exemplar part-based 2D-3D alignment using a large dataset of CAD models

Mathieu Aubry 1, 2, 3, 4 Daniel Maturana 5 Alexei Efros 5, 6 Bryan C. Russell 7 Josef Sivic 4
1 LIGM - Laboratoire d'Informatique Gaspard-Monge
2 ENPC - École des Ponts ParisTech
3 IMAGINE [Marne-la-Vallée]
LIGM - Laboratoire d'Informatique Gaspard-Monge, CSTB - Centre Scientifique et Technique du Bâtiment, ENPC - École des Ponts ParisTech
4 WILLOW - Models of visual object recognition and scene understanding
CNRS - Centre National de la Recherche Scientifique : UMR8548, Inria Paris-Rocquencourt, DI-ENS - Département d'informatique de l'École normale supérieure
5 CMU - Carnegie Mellon University [Pittsburgh]
6 University of California [Berkeley]
7 Intel Research Laboratory
Abstract : This paper poses object category detection in images as a type of 2D-to-3D alignment problem, utilizing the large quantities of 3D CAD models that have been made publicly available online. Using the "chair" class as a running example, we propose an exemplar-based 3D category representation, which can explicitly model chairs of different styles as well as the large variation in viewpoint. We develop an approach to establish part-based correspondences between 3D CAD models and real photographs. This is achieved by (i) representing each 3D model using a set of view-dependent mid-level visual elements learned from synthesized views in a discriminative fashion, (ii) carefully calibrating the individual element detectors on a common dataset of negative images, and (iii) matching visual elements to the test image allowing for small mutual deformations but preserving the viewpoint and style constraints. We demonstrate the ability of our system to align 3D models with 2D objects in the challenging PASCAL VOC images, which depict a wide variety of chairs in complex scenes.
Type de document :
Communication dans un congrès
CVPR 2014 - Compter Vision and Pattern Recognition, Jun 2014, Colombus OH, United States. 2014
Liste complète des métadonnées

Littérature citée [39 références]  Voir  Masquer  Télécharger
https://hal.inria.fr/hal-01057240
Contributeur : Mathieu Aubry <>
Soumis le : jeudi 21 août 2014 - 17:53:02
Dernière modification le : mercredi 30 mai 2018 - 17:22:02
Document(s) archivé(s) le : jeudi 27 novembre 2014 - 13:17:10

Fichier

Aubry14.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01057240, version 1

Collections

INRIA | ENPC | IMAGINE | UPEC-UPEM | PSL | LIGM | PARISTECH

Citation

Mathieu Aubry, Daniel Maturana, Alexei Efros, Bryan C. Russell, Josef Sivic. Seeing 3D chairs: exemplar part-based 2D-3D alignment using a large dataset of CAD models. CVPR 2014 - Compter Vision and Pattern Recognition, Jun 2014, Colombus OH, United States. 2014. 〈hal-01057240〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

603

Téléchargements de fichiers

314

Contact


archive-ouverte@inria.fr