Improving Face Image Representation Using Tangent Vectors and the L1 Norm

Zhicheng LU; Zhizheng LIANG; Lei ZHANG; Jin LIU; Yong ZHOU

doi:10.1587/transfun.E99.A.2099

Improving Face Image Representation Using Tangent Vectors and the L1 Norm

Zhicheng LU, Zhizheng LIANG, Lei ZHANG, Jin LIU, Yong ZHOU

Full Text Views

0

Share
Cite this

Summary :

Inspired from the idea of data representation in manifold learning, we derive a novel model which combines the original training images and their tangent vectors to represent each image in the testing set. Different from the previous methods, the L1 norm is used to control the reconstruction error. Considering the fact that the objective function in the proposed model is non-smooth, we utilize the majorization minimization (MM) method to solve the proposed optimization model. It is interesting to note that at each iteration a quadratic optimization problem is formulated and its analytical solution can be achieved, thereby making the proposed algorithm effective. Extensive experiments on face images demonstrate that our method achieves better performance than some previous methods.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E99-A No.11 pp.2099-2103

Publication Date: 2016/11/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1587/transfun.E99.A.2099

Type of Manuscript: LETTER

Category: Image

Authors

Zhicheng LU
  University of Mining and Technology
Zhizheng LIANG
  University of Mining and Technology
Lei ZHANG
  University of Mining and Technology
Jin LIU
  University of Mining and Technology
Yong ZHOU
  University of Mining and Technology

Keyword

face representation, tangent vectors, majoration minimization methods

Cite this

Copy

Zhicheng LU, Zhizheng LIANG, Lei ZHANG, Jin LIU, Yong ZHOU, "Improving Face Image Representation Using Tangent Vectors and the L1 Norm" in IEICE TRANSACTIONS on Fundamentals, vol. E99-A, no. 11, pp. 2099-2103, November 2016, doi: 10.1587/transfun.E99.A.2099.
Abstract: Inspired from the idea of data representation in manifold learning, we derive a novel model which combines the original training images and their tangent vectors to represent each image in the testing set. Different from the previous methods, the L1 norm is used to control the reconstruction error. Considering the fact that the objective function in the proposed model is non-smooth, we utilize the majorization minimization (MM) method to solve the proposed optimization model. It is interesting to note that at each iteration a quadratic optimization problem is formulated and its analytical solution can be achieved, thereby making the proposed algorithm effective. Extensive experiments on face images demonstrate that our method achieves better performance than some previous methods.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/transfun.E99.A.2099/_p

Copy

@ARTICLE{e99-a_11_2099,
author={Zhicheng LU, Zhizheng LIANG, Lei ZHANG, Jin LIU, Yong ZHOU, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Improving Face Image Representation Using Tangent Vectors and the L1 Norm},
year={2016},
volume={E99-A},
number={11},
pages={2099-2103},
abstract={Inspired from the idea of data representation in manifold learning, we derive a novel model which combines the original training images and their tangent vectors to represent each image in the testing set. Different from the previous methods, the L1 norm is used to control the reconstruction error. Considering the fact that the objective function in the proposed model is non-smooth, we utilize the majorization minimization (MM) method to solve the proposed optimization model. It is interesting to note that at each iteration a quadratic optimization problem is formulated and its analytical solution can be achieved, thereby making the proposed algorithm effective. Extensive experiments on face images demonstrate that our method achieves better performance than some previous methods.},
keywords={},
doi={10.1587/transfun.E99.A.2099},
ISSN={1745-1337},
month={November},}

Copy

TY - JOUR
TI - Improving Face Image Representation Using Tangent Vectors and the L1 Norm
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 2099
EP - 2103
AU - Zhicheng LU
AU - Zhizheng LIANG
AU - Lei ZHANG
AU - Jin LIU
AU - Yong ZHOU
PY - 2016
DO - 10.1587/transfun.E99.A.2099
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E99-A
IS - 11
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - November 2016
AB - Inspired from the idea of data representation in manifold learning, we derive a novel model which combines the original training images and their tangent vectors to represent each image in the testing set. Different from the previous methods, the L1 norm is used to control the reconstruction error. Considering the fact that the objective function in the proposed model is non-smooth, we utilize the majorization minimization (MM) method to solve the proposed optimization model. It is interesting to note that at each iteration a quadratic optimization problem is formulated and its analytical solution can be achieved, thereby making the proposed algorithm effective. Extensive experiments on face images demonstrate that our method achieves better performance than some previous methods.
ER -