A discrete contextual stochastic model for the off-line recognition of handwritten Chinese characters

Author

Xiong, Yan ; Huo, Qiang ; Chan, Chorkin

Author_Institution

Hewlett-Packard Co., Palo Alto, CA, USA

Volume

23

Issue

7

fYear

2001

fDate

7/1/2001 12:00:00 AM

Firstpage

774

Lastpage

782

Abstract

We study a discrete contextual stochastic (CS) model for complex and variant patterns like handwritten Chinese characters. Three fundamental problems of using CS models for character recognition are discussed, and several practical techniques for solving these problems are investigated. A formulation for discriminative training of CS model parameters is also introduced and its practical usage investigated. To illustrate the characteristics of the various algorithms, comparative experiments are performed on a recognition task with a vocabulary consisting of 50 pairs of highly similar handwritten Chinese characters. The experimental results confirm the effectiveness of the discriminative training for improving recognition performance

Keywords

Markov processes; handwritten character recognition; learning (artificial intelligence); pattern matching; Chinese characters; Markov random field; contextual stochastic model; discriminative training; handwritten character recognition; Automatic speech recognition; Character recognition; Context modeling; Handwriting recognition; Hidden Markov models; Pattern matching; Pattern recognition; Shape; Stochastic processes; Vocabulary;

fLanguage

English

Journal_Title

Pattern Analysis and Machine Intelligence, IEEE Transactions on

Publisher

ieee

ISSN

0162-8828

Type

jour

DOI

10.1109/34.935851

Filename

935851