Presentation 2008-02-22
Introduction of Simple Similarity to the Filter Mask Learning in Similar Character Discrimination
Yoshiaki KUROSAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) As the research of the filter mask learning which is the learning system of MCE/GPD (Minimum Classification Error Criterion/Generalized Probabilistic Descent) on the feature extraction, the investigation and the experiments were carried out with the introduction of the inner product for similarity (simple similarity). The results of the method showed that the difference between the inner product and the Euclidean distance was small and both were similar. The filter masks generated by the inner product was also similar to that of Euclidean distance. The case was tested when the normalization and the canonicalization were considered. As the result, a certain improvement was found in the experiment of this case.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) probabilistic descent / OCR / feature extraction / learning / MCE / GPD
Paper # PRMU2007-235
Date of Issue

Conference Information
Committee PRMU
Conference Date 2008/2/14(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) Introduction of Simple Similarity to the Filter Mask Learning in Similar Character Discrimination
Sub Title (in English)
Keyword(1) probabilistic descent
Keyword(2) OCR
Keyword(3) feature extraction
Keyword(4) learning
Keyword(5) MCE
Keyword(6) GPD
1st Author's Name Yoshiaki KUROSAWA
1st Author's Affiliation Toshiba Solustions Corp. Fuchu Engineering Center()
Date 2008-02-22
Paper # PRMU2007-235
Volume (vol) vol.107
Number (no) 491
Page pp.pp.-
#Pages 6
Date of Issue