Handwritten numeral recognition using MFNN based multiexpert combination strategy

Author

Xiaofan Lin ; Ding, Xiaoqing ; Wu, Youshou

Author_Institution

Dept. of Electron. Eng., Tsinghua Univ., Beijing, China

Volume

2

fYear

1997

fDate

18-20 Aug 1997

Firstpage

471

Abstract

A novel unconstrained numeral recognition system using an MFNN (multilayered feedforward neural net)-based multi-expert combination strategy is proposed. Taking advantage of every method´s confidence value as well as its output label, this strategy can improve the system´s performance significantly with as few as two experts. Another novel feature of the system is that the multi-expert combination problem is converted into a new classification problem, and then the MFNNs are used to handle the combination adaptively. The proposed system has achieved promising results on the NIST database

Keywords

adaptive systems; feedforward neural nets; handwriting recognition; image classification; optical character recognition; performance evaluation; NIST database; classification problem; classifier combination; confidence value; handwritten numeral recognition; multi-expert combination strategy; multilayered feedforward neural nets; output label; system performance; unconstrained numeral recognition system; Feedforward neural networks; Feedforward systems; Handwriting recognition; Humans; Image processing; Image recognition; Multi-layer neural network; Neural networks; Pattern recognition; Spatial databases;

fLanguage

English

Publisher

ieee

Conference_Titel

Document Analysis and Recognition, 1997., Proceedings of the Fourth International Conference on

Conference_Location

Ulm

Print_ISBN

0-8186-7898-4

Type

conf

DOI

10.1109/ICDAR.1997.620542

Filename

620542