Presentation 1997/5/15
High Speed and High Accuracy Rough Classification for Handwritten Characters using Hierarchical Learning Vector Quantization
Yuuji Waizumi, Nei Kato, Kazuki Saruta, Yoshiaki Nemoto,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose Hierarchical Learning Vector Quantization(HLVQ) as rough classification method for handwritten characters. HLVQ divides categories in feature space hierarchically in learning procedure. HLVQ adds neurons hierarchically and makes tree whose nodes are neurons. The feature space is divided by a few codebook vectors in each layer. The adjacent feature spaces overlap each other near the borders using window HLVQ possesses both classification speed and accuracy due to the hierarchical architecture and the overlapping technique. By using secondary classification with plural codebook vector for each category under HLVQ, the speed and accuracy of classification by proposed method are higher than that by using average vectors.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) handwritten character recognition / LVQ / Hierarchical LVQ / window
Paper # PRMU97-1
Date of Issue

Conference Information
Committee PRMU
Conference Date 1997/5/15(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) High Speed and High Accuracy Rough Classification for Handwritten Characters using Hierarchical Learning Vector Quantization
Sub Title (in English)
Keyword(1) handwritten character recognition
Keyword(2) LVQ
Keyword(3) Hierarchical LVQ
Keyword(4) window
1st Author's Name Yuuji Waizumi
1st Author's Affiliation Graduate School of Information Sciences, TOHOKU University()
2nd Author's Name Nei Kato
2nd Author's Affiliation Graduate School of Information Sciences, TOHOKU University
3rd Author's Name Kazuki Saruta
3rd Author's Affiliation Faculty of Human Literature and Social Sciences, YAMAGATA University
4th Author's Name Yoshiaki Nemoto
4th Author's Affiliation Graduate School of Information Sciences, TOHOKU University
Date 1997/5/15
Paper # PRMU97-1
Volume (vol) vol.97
Number (no) 40
Page pp.pp.-
#Pages 8
Date of Issue