Presentation 2007-09-03
Class-Distance Weighted Locality Preserving Projection and Its Application To Facial Image Based Age Classification
Masakazu MIYA, Kazuya UEKI, Tetsunori KOBAYASHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A new type of feature transformation technique that extends locality preserving projection (LPP) is proposed and successfully applied to age classification using facial images. We call this technique "class-distance weighted locality preserving projections (CDLPP)". CDLPP performs coordinate transformation using distances between classes as weights while LPP performs coordinate transformation using information on nearby geometric distances between data. In the proposed method, transformation matrices are trained so as to preserve nearby class data structures by weighting the data of nearby classes. An age classification test shows the effectiveness of the proposed method : it reduced the errors of conventional methods such as PCA, LDA, and LPP.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) LPP / CDLPP / Feature Extraction / Facial Image / Age Classification
Paper # PRMU2007-61,HIP2007-70
Date of Issue

Conference Information
Committee HIP
Conference Date 2007/8/27(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 Human Information Processing (HIP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Class-Distance Weighted Locality Preserving Projection and Its Application To Facial Image Based Age Classification
Sub Title (in English)
Keyword(1) LPP
Keyword(2) CDLPP
Keyword(3) Feature Extraction
Keyword(4) Facial Image
Keyword(5) Age Classification
1st Author's Name Masakazu MIYA
1st Author's Affiliation Science and Engineering, Waseda University()
2nd Author's Name Kazuya UEKI
2nd Author's Affiliation NEC Soft, Ltd.
3rd Author's Name Tetsunori KOBAYASHI
3rd Author's Affiliation Science and Engineering, Waseda University
Date 2007-09-03
Paper # PRMU2007-61,HIP2007-70
Volume (vol) vol.107
Number (no) 207
Page pp.pp.-
#Pages 6
Date of Issue