A joint learning based Face Super Resolution approach via contextual topological structure

Liang CHEN; Ruimin Hu; Zhen Han; Zhongyuan Wang; Qing LI

doi:10.1109/ICASSP.2017.7952324

A joint learning based Face Super Resolution approach via contextual topological structure

Liang CHEN
, Ruimin Hu
, Zhen Han
, Zhongyuan Wang
, Qing LI

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

1 Citation (Scopus)

Abstract

Face Super Resolution(FSR) is to infer High Resolution(HR) facial
images from given Low Resolution(LR) ones with the assistance of
LR and HR training pairs. Among existing methods, local patch
based methods are superior in visual and objective quality than global based methods. These local patch based methods are based on
the consistency assumption that the neighbors in HR/LR space form
similar local geometry. But when LR images are with low quality,
the LR space is seriously contaminated that even two distinct patches look similar, which means that the consistency assumption is not well held anymore. To this end, in this paper we introduce the contextual topological structure of target patch to improve the consistency. The contextual topological structure consists of the target patch as well as its adjacent patches, we explore the relationship between them based on statistical probability and apply the relationship for joint learning progress of mapping from LR to HR. By incorporating the contextual topological structure, the robustness to noise of
approach is increased as well as the LR/HR consistency. The effectiveness of proposed method is verified both quantitatively and qualitatively.

Original language	English
Title of host publication	2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Publisher	IEEE
Pages	1088-1092
ISBN (Electronic)	9781509041176
ISBN (Print)	978-1-5090-4118-3
DOIs	https://doi.org/10.1109/ICASSP.2017.7952324
Publication status	Published - Jun 2017
Event	2017 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) - New Orleans, United States Duration: 5 Mar 2017 → 9 Mar 2017 https://ieeexplore.ieee.org/xpl/conhome/7943262/proceeding

Publication series

Name	International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
Publisher	IEEE
ISSN (Print)	1520-6149
ISSN (Electronic)	2379-190X

Conference

Conference	2017 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
Place	United States
City	New Orleans
Period	5/03/17 → 9/03/17
Internet address	https://ieeexplore.ieee.org/xpl/conhome/7943262/proceeding

Research Keywords

consistency enhancement
contextual topological structure
face super resolution
low level vision

Access Output

10.1109/ICASSP.2017.7952324

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

CHEN, L., Hu, R., Han, Z., Wang, Z., & LI, Q. (2017). A joint learning based Face Super Resolution approach via contextual topological structure. In 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 1088-1092). Article 7952324 (International Conference on Acoustics, Speech, and Signal Processing (ICASSP)). IEEE. https://doi.org/10.1109/ICASSP.2017.7952324

@inproceedings{1b940a9e51824fefa04a0f4c79ab8c53,

title = "A joint learning based Face Super Resolution approach via contextual topological structure",

abstract = "Face Super Resolution(FSR) is to infer High Resolution(HR) facial images from given Low Resolution(LR) ones with the assistance of LR and HR training pairs. Among existing methods, local patch based methods are superior in visual and objective quality than global based methods. These local patch based methods are based on the consistency assumption that the neighbors in HR/LR space form similar local geometry. But when LR images are with low quality, the LR space is seriously contaminated that even two distinct patches look similar, which means that the consistency assumption is not well held anymore. To this end, in this paper we introduce the contextual topological structure of target patch to improve the consistency. The contextual topological structure consists of the target patch as well as its adjacent patches, we explore the relationship between them based on statistical probability and apply the relationship for joint learning progress of mapping from LR to HR. By incorporating the contextual topological structure, the robustness to noise of approach is increased as well as the LR/HR consistency. The effectiveness of proposed method is verified both quantitatively and qualitatively. ",

keywords = "consistency enhancement, contextual topological structure, face super resolution, low level vision",

author = "Liang CHEN and Ruimin Hu and Zhen Han and Zhongyuan Wang and Qing LI",

year = "2017",

month = jun,

doi = "10.1109/ICASSP.2017.7952324",

language = "English",

isbn = "978-1-5090-4118-3",

series = "International Conference on Acoustics, Speech, and Signal Processing (ICASSP)",

publisher = "IEEE",

pages = "1088--1092",

booktitle = "2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)",

address = "United States",

note = "2017 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) ; Conference date: 05-03-2017 Through 09-03-2017",

url = "https://ieeexplore.ieee.org/xpl/conhome/7943262/proceeding",

}

CHEN, L, Hu, R, Han, Z, Wang, Z & LI, Q 2017, A joint learning based Face Super Resolution approach via contextual topological structure. in 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)., 7952324, International Conference on Acoustics, Speech, and Signal Processing (ICASSP), IEEE, pp. 1088-1092, 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), New Orleans, Louisiana, United States, 5/03/17. https://doi.org/10.1109/ICASSP.2017.7952324

A joint learning based Face Super Resolution approach via contextual topological structure. / CHEN, Liang; Hu, Ruimin; Han, Zhen et al.
2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2017. p. 1088-1092 7952324 (International Conference on Acoustics, Speech, and Signal Processing (ICASSP)).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - A joint learning based Face Super Resolution approach via contextual topological structure

AU - CHEN, Liang

AU - Hu, Ruimin

AU - Han, Zhen

AU - Wang, Zhongyuan

AU - LI, Qing

PY - 2017/6

Y1 - 2017/6

N2 - Face Super Resolution(FSR) is to infer High Resolution(HR) facial images from given Low Resolution(LR) ones with the assistance of LR and HR training pairs. Among existing methods, local patch based methods are superior in visual and objective quality than global based methods. These local patch based methods are based on the consistency assumption that the neighbors in HR/LR space form similar local geometry. But when LR images are with low quality, the LR space is seriously contaminated that even two distinct patches look similar, which means that the consistency assumption is not well held anymore. To this end, in this paper we introduce the contextual topological structure of target patch to improve the consistency. The contextual topological structure consists of the target patch as well as its adjacent patches, we explore the relationship between them based on statistical probability and apply the relationship for joint learning progress of mapping from LR to HR. By incorporating the contextual topological structure, the robustness to noise of approach is increased as well as the LR/HR consistency. The effectiveness of proposed method is verified both quantitatively and qualitatively.

AB - Face Super Resolution(FSR) is to infer High Resolution(HR) facial images from given Low Resolution(LR) ones with the assistance of LR and HR training pairs. Among existing methods, local patch based methods are superior in visual and objective quality than global based methods. These local patch based methods are based on the consistency assumption that the neighbors in HR/LR space form similar local geometry. But when LR images are with low quality, the LR space is seriously contaminated that even two distinct patches look similar, which means that the consistency assumption is not well held anymore. To this end, in this paper we introduce the contextual topological structure of target patch to improve the consistency. The contextual topological structure consists of the target patch as well as its adjacent patches, we explore the relationship between them based on statistical probability and apply the relationship for joint learning progress of mapping from LR to HR. By incorporating the contextual topological structure, the robustness to noise of approach is increased as well as the LR/HR consistency. The effectiveness of proposed method is verified both quantitatively and qualitatively.

KW - consistency enhancement

KW - contextual topological structure

KW - face super resolution

KW - low level vision

UR - https://www.scopus.com/pages/publications/85023758390#tab=impact

UR - https://www.scopus.com/pages/publications/85023758390

U2 - 10.1109/ICASSP.2017.7952324

DO - 10.1109/ICASSP.2017.7952324

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 978-1-5090-4118-3

T3 - International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

SP - 1088

EP - 1092

BT - 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

PB - IEEE

T2 - 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

Y2 - 5 March 2017 through 9 March 2017

ER -

A joint learning based Face Super Resolution approach via contextual topological structure

Abstract

Publication series

Conference

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this