Facial Image Super-Resolution Reconstruction Based on Separated Frequency Components

Hyunduk KIM; Sang-Heon LEE; Myoung-Kyu SOHN; Dong-Ju KIM; Byungmin KIM

doi:10.1587/transfun.E96.A.1315

Facial Image Super-Resolution Reconstruction Based on Separated Frequency Components

Hyunduk KIM, Sang-Heon LEE, Myoung-Kyu SOHN, Dong-Ju KIM, Byungmin KIM

Full Text Views

0

Share
Cite this

Summary :

Super resolution (SR) reconstruction is the process of fusing a sequence of low-resolution images into one high-resolution image. Many researchers have introduced various SR reconstruction methods. However, these traditional methods are limited in the extent to which they allow recovery of high-frequency information. Moreover, due to the self-similarity of face images, most of the facial SR algorithms are machine learning based. In this paper, we introduce a facial SR algorithm that combines learning-based and regularized SR image reconstruction algorithms. Our conception involves two main ideas. First, we employ separated frequency components to reconstruct high-resolution images. In addition, we separate the region of the training face image. These approaches can help to recover high-frequency information. In our experiments, we demonstrate the effectiveness of these ideas.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E96-A No.6 pp.1315-1322

Publication Date: 2013/06/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1587/transfun.E96.A.1315

Type of Manuscript: Special Section PAPER (Special Section on Circuit, System, and Computer Technologies)

Category

Authors

Hyunduk KIM
  Daegu Gyeongbuk Institute of Science & Technology (DGIST)
Sang-Heon LEE
  Daegu Gyeongbuk Institute of Science & Technology (DGIST)
Myoung-Kyu SOHN
  Daegu Gyeongbuk Institute of Science & Technology (DGIST)
Dong-Ju KIM
  Daegu Gyeongbuk Institute of Science & Technology (DGIST)
Byungmin KIM
  Daegu Gyeongbuk Institute of Science & Technology (DGIST)

Keyword

facial super resolution, frequency domain, sparse representation, regularization technique, bilateral filter

Cite this

Copy

Hyunduk KIM, Sang-Heon LEE, Myoung-Kyu SOHN, Dong-Ju KIM, Byungmin KIM, "Facial Image Super-Resolution Reconstruction Based on Separated Frequency Components" in IEICE TRANSACTIONS on Fundamentals, vol. E96-A, no. 6, pp. 1315-1322, June 2013, doi: 10.1587/transfun.E96.A.1315.
Abstract: Super resolution (SR) reconstruction is the process of fusing a sequence of low-resolution images into one high-resolution image. Many researchers have introduced various SR reconstruction methods. However, these traditional methods are limited in the extent to which they allow recovery of high-frequency information. Moreover, due to the self-similarity of face images, most of the facial SR algorithms are machine learning based. In this paper, we introduce a facial SR algorithm that combines learning-based and regularized SR image reconstruction algorithms. Our conception involves two main ideas. First, we employ separated frequency components to reconstruct high-resolution images. In addition, we separate the region of the training face image. These approaches can help to recover high-frequency information. In our experiments, we demonstrate the effectiveness of these ideas.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/transfun.E96.A.1315/_p

Copy

@ARTICLE{e96-a_6_1315,
author={Hyunduk KIM, Sang-Heon LEE, Myoung-Kyu SOHN, Dong-Ju KIM, Byungmin KIM, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Facial Image Super-Resolution Reconstruction Based on Separated Frequency Components},
year={2013},
volume={E96-A},
number={6},
pages={1315-1322},
abstract={Super resolution (SR) reconstruction is the process of fusing a sequence of low-resolution images into one high-resolution image. Many researchers have introduced various SR reconstruction methods. However, these traditional methods are limited in the extent to which they allow recovery of high-frequency information. Moreover, due to the self-similarity of face images, most of the facial SR algorithms are machine learning based. In this paper, we introduce a facial SR algorithm that combines learning-based and regularized SR image reconstruction algorithms. Our conception involves two main ideas. First, we employ separated frequency components to reconstruct high-resolution images. In addition, we separate the region of the training face image. These approaches can help to recover high-frequency information. In our experiments, we demonstrate the effectiveness of these ideas.},
keywords={},
doi={10.1587/transfun.E96.A.1315},
ISSN={1745-1337},
month={June},}

Copy

TY - JOUR
TI - Facial Image Super-Resolution Reconstruction Based on Separated Frequency Components
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1315
EP - 1322
AU - Hyunduk KIM
AU - Sang-Heon LEE
AU - Myoung-Kyu SOHN
AU - Dong-Ju KIM
AU - Byungmin KIM
PY - 2013
DO - 10.1587/transfun.E96.A.1315
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E96-A
IS - 6
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - June 2013
AB - Super resolution (SR) reconstruction is the process of fusing a sequence of low-resolution images into one high-resolution image. Many researchers have introduced various SR reconstruction methods. However, these traditional methods are limited in the extent to which they allow recovery of high-frequency information. Moreover, due to the self-similarity of face images, most of the facial SR algorithms are machine learning based. In this paper, we introduce a facial SR algorithm that combines learning-based and regularized SR image reconstruction algorithms. Our conception involves two main ideas. First, we employ separated frequency components to reconstruct high-resolution images. In addition, we separate the region of the training face image. These approaches can help to recover high-frequency information. In our experiments, we demonstrate the effectiveness of these ideas.
ER -