Création et utilisation de vocabulaires visuels pour la catégorisation d'images et la segmentation de classes d'objets

Diane Larlus 1, 2
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : This thesis deals with the interpretation of static images, with a focus on recognising object categories. We consider several different approaches, which are all variations on the bag-of-words model, and all use local image descriptors. Local descriptors are quantised using a visual vocabulary.
The first part of the thesis examines different methods for creating visual vocabularies. We aim to create vocabularies which perform well for image categorisation.
The second part of the thesis deals with object category segmentation. In particular, we present a method which combines the low-level consistency properties of a Markov Random Field with an appearance model which provides higher-level constraints. The appearance model is based on regions which each represent a single object as a set of visual words.
Document type :
Theses
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00343665
Contributor : Diane Larlus <>
Submitted on : Tuesday, December 2, 2008 - 2:51:55 PM
Last modification on : Monday, December 17, 2018 - 11:22:02 AM
Long-term archiving on : Thursday, October 11, 2012 - 12:25:10 PM

Identifiers

  • HAL Id : tel-00343665, version 1

Collections

Citation

Diane Larlus. Création et utilisation de vocabulaires visuels pour la catégorisation d'images et la segmentation de classes d'objets. Interface homme-machine [cs.HC]. Institut National Polytechnique de Grenoble - INPG, 2008. Français. ⟨tel-00343665⟩

Share

Metrics

Record views

786

Files downloads

647