Maximum Volume Constrained Graph Nonnegative Matrix Factorization for Facial Expression Recognition

Viet-Hang DUONG; Manh-Quan BUI; Jian-Jiun DING; Bach-Tung PHAM; Pham The BAO; Jia-Ching WANG

doi:10.1587/transfun.E100.A.3081

Maximum Volume Constrained Graph Nonnegative Matrix Factorization for Facial Expression Recognition

Viet-Hang DUONG, Manh-Quan BUI, Jian-Jiun DING, Bach-Tung PHAM, Pham The BAO, Jia-Ching WANG

Full Text Views

0

Share
Cite this

Summary :

In this work, two new proposed NMF models are developed for facial expression recognition. They are called maximum volume constrained nonnegative matrix factorization (MV_NMF) and maximum volume constrained graph nonnegative matrix factorization (MV_GNMF). They achieve sparseness from a larger simplicial cone constraint and the extracted features preserve the topological structure of the original images.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E100-A No.12 pp.3081-3085

Publication Date: 2017/12/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1587/transfun.E100.A.3081

Type of Manuscript: LETTER

Category: Image

Authors

Viet-Hang DUONG
  National Central University
Manh-Quan BUI
  National Central University
Jian-Jiun DING
  National Taiwan University
Bach-Tung PHAM
  National Central University
Pham The BAO
  University of Science
Jia-Ching WANG
  National Central University

Keyword

facial expression recognition, nonnegative matrix factorization, feature extraction, graph regularized, projected gradient descent

Cite this

Copy

Viet-Hang DUONG, Manh-Quan BUI, Jian-Jiun DING, Bach-Tung PHAM, Pham The BAO, Jia-Ching WANG, "Maximum Volume Constrained Graph Nonnegative Matrix Factorization for Facial Expression Recognition" in IEICE TRANSACTIONS on Fundamentals, vol. E100-A, no. 12, pp. 3081-3085, December 2017, doi: 10.1587/transfun.E100.A.3081.
Abstract: In this work, two new proposed NMF models are developed for facial expression recognition. They are called maximum volume constrained nonnegative matrix factorization (MV_NMF) and maximum volume constrained graph nonnegative matrix factorization (MV_GNMF). They achieve sparseness from a larger simplicial cone constraint and the extracted features preserve the topological structure of the original images.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/transfun.E100.A.3081/_p

Copy

@ARTICLE{e100-a_12_3081,
author={Viet-Hang DUONG, Manh-Quan BUI, Jian-Jiun DING, Bach-Tung PHAM, Pham The BAO, Jia-Ching WANG, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Maximum Volume Constrained Graph Nonnegative Matrix Factorization for Facial Expression Recognition},
year={2017},
volume={E100-A},
number={12},
pages={3081-3085},
abstract={In this work, two new proposed NMF models are developed for facial expression recognition. They are called maximum volume constrained nonnegative matrix factorization (MV_NMF) and maximum volume constrained graph nonnegative matrix factorization (MV_GNMF). They achieve sparseness from a larger simplicial cone constraint and the extracted features preserve the topological structure of the original images.},
keywords={},
doi={10.1587/transfun.E100.A.3081},
ISSN={1745-1337},
month={December},}

Copy

TY - JOUR
TI - Maximum Volume Constrained Graph Nonnegative Matrix Factorization for Facial Expression Recognition
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 3081
EP - 3085
AU - Viet-Hang DUONG
AU - Manh-Quan BUI
AU - Jian-Jiun DING
AU - Bach-Tung PHAM
AU - Pham The BAO
AU - Jia-Ching WANG
PY - 2017
DO - 10.1587/transfun.E100.A.3081
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E100-A
IS - 12
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - December 2017
AB - In this work, two new proposed NMF models are developed for facial expression recognition. They are called maximum volume constrained nonnegative matrix factorization (MV_NMF) and maximum volume constrained graph nonnegative matrix factorization (MV_GNMF). They achieve sparseness from a larger simplicial cone constraint and the extracted features preserve the topological structure of the original images.
ER -