Presentation 2008-02-21
Facial Expression Recognition Using Fisher Weight Maps for Gabor Features
Tomohisa SHIMOTSU, Takashi TAKAHASHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Fisher Weight Maps method is one of feature extraction methods for image-based pattern recognition. It applies linear discriminant analysis to a set of data matrices composed of local image features, and finds the optimal weight matrix which extracts discriminative features from the data matrices. We investigate two modifications on this method: 1) employing Gabor filtering to extract local features instead of Higher-order Local Auto-Correlation (HLAC), 2) composing the weights for each type of local features (corresponding to scales and orientations of Gabor filters ) instead of the weights for each position. It is shown that the proposed method outperforms the original method on two recognition tasks: facial expression recognition and pedestrian/non-pedestrian recognition.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Fisher Weight Maps / Facial Expression Recognition / Linear Discriminant Analysis / Gabor Features
Paper # PRMU2007-221
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) Facial Expression Recognition Using Fisher Weight Maps for Gabor Features
Sub Title (in English)
Keyword(1) Fisher Weight Maps
Keyword(2) Facial Expression Recognition
Keyword(3) Linear Discriminant Analysis
Keyword(4) Gabor Features
1st Author's Name Tomohisa SHIMOTSU
1st Author's Affiliation Graduate School of Science and Technology, Ryukoku University()
2nd Author's Name Takashi TAKAHASHI
2nd Author's Affiliation Faculty of Science and Technology, Ryukoku University
Date 2008-02-21
Paper # PRMU2007-221
Volume (vol) vol.107
Number (no) 491
Page pp.pp.-
#Pages 6
Date of Issue