DSVD: A tensor-based image compression and recognition method

Kohei Inoue, Kiichi Urahama

Research output: Contribution to journalConference article

12 Citations (Scopus)

Abstract

Optimal dimensionality reduction of a single matrix is given by the truncated singular value decomposition, and optimal compression of a set of vector data is given by the principal component analysis. We present, in this paper, a dyadic singular value decomposition (DSVD) which gives a near-optimal dimensionality reduction of a set of matrix data and apply it to image compression and face recognition. The DSVD algorithm is derived from the higher-order singular value decomposition (HOSVD) of a third-order tensor and gives an analytical solution of a low-rank approximation problem for data matrices. The DSVD outperforms the other dimensionality reduction methods in the computational speed and accuracy in image compression. Its face recognition rate is higher than the eigenface method.

Original languageEnglish
Article number1466083
Pages (from-to)6308-6311
Number of pages4
JournalProceedings - IEEE International Symposium on Circuits and Systems
DOIs
Publication statusPublished - Dec 1 2005
EventIEEE International Symposium on Circuits and Systems 2005, ISCAS 2005 - Kobe, Japan
Duration: May 23 2005May 26 2005

Fingerprint

Image recognition
Singular value decomposition
Image compression
Tensors
compression
decomposition
Face recognition
matrix
Principal component analysis
principal component analysis
method

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

DSVD : A tensor-based image compression and recognition method. / Inoue, Kohei; Urahama, Kiichi.

In: Proceedings - IEEE International Symposium on Circuits and Systems, 01.12.2005, p. 6308-6311.

Research output: Contribution to journalConference article

@article{83580b6605524671bc826e74b2ec68e5,
title = "DSVD: A tensor-based image compression and recognition method",
abstract = "Optimal dimensionality reduction of a single matrix is given by the truncated singular value decomposition, and optimal compression of a set of vector data is given by the principal component analysis. We present, in this paper, a dyadic singular value decomposition (DSVD) which gives a near-optimal dimensionality reduction of a set of matrix data and apply it to image compression and face recognition. The DSVD algorithm is derived from the higher-order singular value decomposition (HOSVD) of a third-order tensor and gives an analytical solution of a low-rank approximation problem for data matrices. The DSVD outperforms the other dimensionality reduction methods in the computational speed and accuracy in image compression. Its face recognition rate is higher than the eigenface method.",
author = "Kohei Inoue and Kiichi Urahama",
year = "2005",
month = "12",
day = "1",
doi = "10.1109/ISCAS.2005.1466083",
language = "English",
pages = "6308--6311",
journal = "Proceedings - IEEE International Symposium on Circuits and Systems",
issn = "0271-4310",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - DSVD

T2 - A tensor-based image compression and recognition method

AU - Inoue, Kohei

AU - Urahama, Kiichi

PY - 2005/12/1

Y1 - 2005/12/1

N2 - Optimal dimensionality reduction of a single matrix is given by the truncated singular value decomposition, and optimal compression of a set of vector data is given by the principal component analysis. We present, in this paper, a dyadic singular value decomposition (DSVD) which gives a near-optimal dimensionality reduction of a set of matrix data and apply it to image compression and face recognition. The DSVD algorithm is derived from the higher-order singular value decomposition (HOSVD) of a third-order tensor and gives an analytical solution of a low-rank approximation problem for data matrices. The DSVD outperforms the other dimensionality reduction methods in the computational speed and accuracy in image compression. Its face recognition rate is higher than the eigenface method.

AB - Optimal dimensionality reduction of a single matrix is given by the truncated singular value decomposition, and optimal compression of a set of vector data is given by the principal component analysis. We present, in this paper, a dyadic singular value decomposition (DSVD) which gives a near-optimal dimensionality reduction of a set of matrix data and apply it to image compression and face recognition. The DSVD algorithm is derived from the higher-order singular value decomposition (HOSVD) of a third-order tensor and gives an analytical solution of a low-rank approximation problem for data matrices. The DSVD outperforms the other dimensionality reduction methods in the computational speed and accuracy in image compression. Its face recognition rate is higher than the eigenface method.

UR - http://www.scopus.com/inward/record.url?scp=33845562150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845562150&partnerID=8YFLogxK

U2 - 10.1109/ISCAS.2005.1466083

DO - 10.1109/ISCAS.2005.1466083

M3 - Conference article

AN - SCOPUS:33845562150

SP - 6308

EP - 6311

JO - Proceedings - IEEE International Symposium on Circuits and Systems

JF - Proceedings - IEEE International Symposium on Circuits and Systems

SN - 0271-4310

M1 - 1466083

ER -