Share Email Print
cover

Proceedings Paper

Construction of language models for an handwritten mail reading system
Author(s): Olivier Morillot; Laurence Likforman-Sulem; Emmanuèle Grosicki
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

This paper presents a system for the recognition of unconstrained handwritten mails. The main part of this system is an HMM recognizer which uses trigraphs to model contextual information. This recognition system does not require any segmentation into words or characters and directly works at line level. To take into account linguistic information and enhance performance, a language model is introduced. This language model is based on bigrams and built from training document transcriptions only. Different experiments with various vocabulary sizes and language models have been conducted. Word Error Rate and Perplexity values are compared to show the interest of specific language models, fit to handwritten mail recognition task.

Paper Details

Date Published: 24 January 2012
PDF: 9 pages
Proc. SPIE 8297, Document Recognition and Retrieval XIX, 82970S (24 January 2012); doi: 10.1117/12.911965
Show Author Affiliations
Olivier Morillot, Lab. Traitement et Communication de l'Information, CNRS, Telecom ParisTech (France)
Laurence Likforman-Sulem, Lab. Traitement et Communication de l'Information, CNRS, Telecom ParisTech (France)
Emmanuèle Grosicki, DGA (France)


Published in SPIE Proceedings Vol. 8297:
Document Recognition and Retrieval XIX
Christian Viard-Gaudin; Richard Zanibbi, Editor(s)

© SPIE. Terms of Use
Back to Top