Scene Text Recognition and Retrieval for Large Lexicons

Udit Roy 1 Anand Mishra 1 Karteek Alahari 2 C.V. Jawahar 1
2 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : In this paper we propose a framework for recognition and retrieval tasks in the context of scene text images. In contrast to many of the recent works, we focus on the case where an image-specific list of words, known as the small lexicon setting, is unavailable. We present a conditional random field model defined on potential character locations and the interactions between them. Observing that the interaction potentials computed in the large lexicon setting are less effective than in the case of a small lexicon, we propose an iterative method, which alternates between finding the most likely solution and refining the interaction po-tentials. We evaluate our method on public datasets and show that it improves over baseline and state-of-the-art approaches. For example, we obtain nearly 15% improvement in recognition accuracy and precision for our retrieval task over baseline methods on the IIIT-5K word dataset, with a large lexicon containing 0.5 million words.
Type de document :
Communication dans un congrès
ACCV - Asian Conference on Computer Vision, Nov 2014, Singapore, Singapore. <http://www.accv2014.org>
Liste complète des métadonnées


https://hal.inria.fr/hal-01088739
Contributeur : Karteek Alahari <>
Soumis le : vendredi 28 novembre 2014 - 15:26:13
Dernière modification le : vendredi 24 avril 2015 - 01:01:32
Document(s) archivé(s) le : vendredi 14 avril 2017 - 23:03:30

Fichier

roy14.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01088739, version 1

Collections

Citation

Udit Roy, Anand Mishra, Karteek Alahari, C.V. Jawahar. Scene Text Recognition and Retrieval for Large Lexicons. ACCV - Asian Conference on Computer Vision, Nov 2014, Singapore, Singapore. <http://www.accv2014.org>. <hal-01088739>

Partager

Métriques

Consultations de
la notice

337

Téléchargements du document

346