Maximum mutual information training for an online neural predictive handwritten word recognition system |
| |
Authors: | Sonia Garcia-Salicetti Bernadette Dorizzi Patrick Gallinari Zsolt Wimmer |
| |
Affiliation: | (1) Laboratoire de Reconnaissance des Formes et Vision (RFV), Institut National des Sciences Appliquées de Lyon, Bat. 403, 20 Avenue Albert Einstein, 69621 Villeurbanne, France , FR;(2) Département Electronique et Physique (EPH), Institut National des Télécommunications, 9 rue Charles Fourier, 91011 Evry, France , FR;(3) Laboratoire d'Informatique de Paris 6 (LIP6), Université Paris 6, 4 Place Jussieu, 75252 Paris 05, France , FR;(4) Société Vision Objects, 11 rue de la Fontaine Caron, 44300 Nantes, France , FR |
| |
Abstract: | In this paper, we present a hybrid online handwriting recognition system based on hidden Markov models (HMMs). It is devoted to word recognition using large vocabularies. An adaptive segmentation of words into letters is integrated with recognition, and is at the heart of the training phase. A word-model is a left-right HMM in which each state is a predictive multilayer perceptron that performs local regression on the drawing (i.e., the written word) relying on a context of observations. A discriminative training paradigm related to maximum mutual information is used, and its potential is shown on a database of 9,781 words. Received June 19, 2000 / Revised October 16, 2000 |
| |
Keywords: | : Predictive neural network – Hidden Markov model – Maximum mutual information – Dynamic segmentation – Handwritten word recognition |
本文献已被 SpringerLink 等数据库收录! |
|