Presentation 2006-03-16
Hierarchical Clustering Of Feature Vectors at Visual Attentional Points
Jun SAITO, Hayato YAMANA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In Content-based image retrieval, the classifications is needed for better performances of (i) speeds of retrieval and (ii) semanticity of retrieval. Our system extracts the most attentional points by using a selective visual attention model which extracts feature vectors of attentional points in images. And our system classifies feature vectors by hierarchical clustering with residuals. An attentional point in an image is outlier in an image, or special outlier. We propose extension of selective attention model to extract temporal outlier with residual vectors, and the method of moving attentional points weighted by a cancroid of a category extracted. This paper shows that the outlier extraction idea of selective visual attention is extensible to extract hierarchical categories. And also, this paper shows that our method can select an image point which belongs to an extracted category.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Selective Visual Attention / Content-based Image Retrieval / Hierarchical Clustering With Residual
Paper # PRMU2005-241
Date of Issue

Conference Information
Committee PRMU
Conference Date 2006/3/9(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) Hierarchical Clustering Of Feature Vectors at Visual Attentional Points
Sub Title (in English)
Keyword(1) Selective Visual Attention
Keyword(2) Content-based Image Retrieval
Keyword(3) Hierarchical Clustering With Residual
1st Author's Name Jun SAITO
1st Author's Affiliation ()
2nd Author's Name Hayato YAMANA
2nd Author's Affiliation Science and Engineering Waseda, University Okubo:National Institute of Informatics Hitotsubashi
Date 2006-03-16
Paper # PRMU2005-241
Volume (vol) vol.105
Number (no) 673
Page pp.pp.-
#Pages 6
Date of Issue