Cross-learning in Analytic Word Recognition Without Segmentation

Christophe Choisy 1 Abdel Belaïd 1
1 READ - READ
LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications
Abstract : In this paper a method for analytic handwritten word recognition based on causal Markov random fields is described. The word models are hmms where each state corresponds to a letter; each letter is modeled by a nshp (Markov field). The word models are built dynamically. Letter and word model training is made using baum algorithm where the parameters are re-estimated on the generated word global models. The segmentation is not necessary~: the system determines itself during training the best repartition of the information within the letter models. First experiments on two real databases of french check amount words give very encouraging results up to 86% for recognition.
Type de document :
Article dans une revue
International Journal on Document Analysis and Recognition, Springer Verlag, 2002, 4 (4), pp.281-289
Liste complète des métadonnées

https://hal.inria.fr/inria-00100851
Contributeur : Publications Loria <>
Soumis le : mardi 26 septembre 2006 - 14:52:25
Dernière modification le : mardi 24 avril 2018 - 13:34:29

Identifiants

  • HAL Id : inria-00100851, version 1

Collections

Citation

Christophe Choisy, Abdel Belaïd. Cross-learning in Analytic Word Recognition Without Segmentation. International Journal on Document Analysis and Recognition, Springer Verlag, 2002, 4 (4), pp.281-289. 〈inria-00100851〉

Partager

Métriques

Consultations de la notice

94