ORTHOGONAL NONNEGATIVE TUCKER DECOMPOSITION

Junjun PAN; Michael K. NG; Ye LIU; Xiongjun ZHANG; Hong YAN

doi:10.1137/19M1294708

ORTHOGONAL NONNEGATIVE TUCKER DECOMPOSITION

Junjun PAN
, Michael K. NG
, Ye LIU
, Xiongjun ZHANG
, Hong YAN

Department of Electrical Engineering

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

34 Citations (Scopus)

120 Downloads (CityUHK Scholars)

Abstract

In this paper, we study nonnegative tensor data and propose an orthogonal nonnegative Tucker decomposition (ONTD). We discuss some properties of ONTD and develop a convex relaxation algorithm of the augmented Lagrangian function to solve the optimization problem. The convergence of the algorithm is given. We employ ONTD on the image data sets from the real world applications including face recognition, image representation, and hyperspectral unmixing. Numerical results are shown to illustrate the effectiveness of the proposed algorithm.

Original language	English
Pages (from-to)	B55-B81
Journal	SIAM Journal on Scientific Computing
Volume	43
Issue number	1
Online published	7 Jan 2021
DOIs	https://doi.org/10.1137/19M1294708
Publication status	Published - 2021

Research Keywords

Image processing
Nonnegative tensor
Tucker decomposition

Publisher's Copyright Statement

COPYRIGHT TERMS OF DEPOSITED FINAL PUBLISHED VERSION FILE: © 2021 Society for Industrial and Applied Mathematics.

RGC Funding Information

RGC-funded

Access Output

10.1137/19M1294708

73625976.pdfFinal Published version, 949 KB

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{69655a0e1eac4c5187bc814abee6931c,

title = "ORTHOGONAL NONNEGATIVE TUCKER DECOMPOSITION",

abstract = "In this paper, we study nonnegative tensor data and propose an orthogonal nonnegative Tucker decomposition (ONTD). We discuss some properties of ONTD and develop a convex relaxation algorithm of the augmented Lagrangian function to solve the optimization problem. The convergence of the algorithm is given. We employ ONTD on the image data sets from the real world applications including face recognition, image representation, and hyperspectral unmixing. Numerical results are shown to illustrate the effectiveness of the proposed algorithm.",

keywords = "Image processing, Nonnegative tensor, Tucker decomposition",

author = "Junjun PAN and NG, \{Michael K.\} and Ye LIU and Xiongjun ZHANG and Hong YAN",

year = "2021",

doi = "10.1137/19M1294708",

language = "English",

volume = "43",

pages = "B55--B81",

journal = "SIAM Journal on Scientific Computing",

issn = "1064-8275",

publisher = "Society for Industrial and Applied Mathematics",

number = "1",

}

TY - JOUR

T1 - ORTHOGONAL NONNEGATIVE TUCKER DECOMPOSITION

AU - PAN, Junjun

AU - NG, Michael K.

AU - LIU, Ye

AU - ZHANG, Xiongjun

AU - YAN, Hong

PY - 2021

Y1 - 2021

N2 - In this paper, we study nonnegative tensor data and propose an orthogonal nonnegative Tucker decomposition (ONTD). We discuss some properties of ONTD and develop a convex relaxation algorithm of the augmented Lagrangian function to solve the optimization problem. The convergence of the algorithm is given. We employ ONTD on the image data sets from the real world applications including face recognition, image representation, and hyperspectral unmixing. Numerical results are shown to illustrate the effectiveness of the proposed algorithm.

AB - In this paper, we study nonnegative tensor data and propose an orthogonal nonnegative Tucker decomposition (ONTD). We discuss some properties of ONTD and develop a convex relaxation algorithm of the augmented Lagrangian function to solve the optimization problem. The convergence of the algorithm is given. We employ ONTD on the image data sets from the real world applications including face recognition, image representation, and hyperspectral unmixing. Numerical results are shown to illustrate the effectiveness of the proposed algorithm.

KW - Image processing

KW - Nonnegative tensor

KW - Tucker decomposition

UR - https://www.scopus.com/pages/publications/85102809410

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85102809410&origin=recordpage

U2 - 10.1137/19M1294708

DO - 10.1137/19M1294708

M3 - RGC 21 - Publication in refereed journal

SN - 1064-8275

VL - 43

SP - B55-B81

JO - SIAM Journal on Scientific Computing

JF - SIAM Journal on Scientific Computing

IS - 1

ER -

ORTHOGONAL NONNEGATIVE TUCKER DECOMPOSITION

Abstract

Research Keywords

Publisher's Copyright Statement

RGC Funding Information

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

CRF: Efficient Algorithms and Hardware Accelerators for Tensor Decomposition and Their Applications to Multidimensional Data Analysis

Cite this

ORTHOGONAL NONNEGATIVE TUCKER DECOMPOSITION

Abstract

Research Keywords

Publisher's Copyright Statement

RGC Funding Information

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Projects

CRF: Efficient Algorithms and Hardware Accelerators for Tensor Decomposition and Their Applications to Multidimensional Data Analysis

Cite this