A Study of Designing Compact Recognizers of Handwritten Chinese Characters Using Multiple-Prototype Based Classifiers

Author

Wang, Yongqiang ; Huo, Qiang

Author_Institution

Microsoft Res. Asia, Beijing, China

fYear

2010

fDate

23-26 Aug. 2010

Firstpage

1872

Lastpage

1875

Abstract

We present a study of designing compact recognizers of handwritten Chinese characters using multiple-prototype based classifiers. A modified Quick prop algorithm is proposed to optimize a sample-separation-margin based minimum classification error objective function. Split vector quantization technique is used to compress classifier parameters. Benchmark results are reported for classifiers with different footprints trained from about 10 million samples on a recognition task with a vocabulary of 9282 character classes which include 9119 Chinese characters, 62 alphanumeric characters, 101 punctuation marks and symbols.

Keywords

handwriting recognition; natural language processing; pattern classification; vector quantisation; compact recognizers design; handwritten Chinese characters; minimum classification error objective function; modified Quickprop algorithm; multiple prototype based classifiers; sample separation margin; split vector quantization technique; Accuracy; Character recognition; Feature extraction; Handwriting recognition; Prototypes; Training; Vocabulary; handwriting recognition; large margin; minimum classification error; pattern recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition (ICPR), 2010 20th International Conference on

Conference_Location

Istanbul

ISSN

1051-4651

Print_ISBN

978-1-4244-7542-1

Type

conf

DOI

10.1109/ICPR.2010.1138

Filename

5597229