Generative Adversarial Networks-based Face Hallucination with Identity-Preserving

Chia Hung Yeh; Daniel Chiu; Li Wei Kang; Chih Chung Hsu; Chen Lo

doi:10.1109/ICCE-TW52618.2021.9603171

Generative Adversarial Networks-based Face Hallucination with Identity-Preserving

Chia Hung Yeh, Daniel Chiu, Li Wei Kang, Chih Chung Hsu, Chen Lo

Department of Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This paper presents a novel generative adversarial networks-based face hallucination framework for producing high-resolution face images from very low-resolution (LR) ones. We propose a multi-scale generator architecture with multi-scale loss functions for different upscaling factors and a triplet-based identity preserving loss for extracting multi-scale identity-aware facial representations. Experimental results have verified that our method can well super-resolve very LR face images (e.g., 8×8) quantitatively and qualitatively.

Original language	English
Title of host publication	2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781665433280
DOIs	https://doi.org/10.1109/ICCE-TW52618.2021.9603171
Publication status	Published - 2021
Event	8th IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021 - Penghu, Taiwan Duration: 2021 Sept 15 → 2021 Sept 17

Publication series

Name	2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021

Conference

Conference	8th IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021
Country/Territory	Taiwan
City	Penghu
Period	2021/09/15 → 2021/09/17

ASJC Scopus subject areas

Artificial Intelligence
Computer Science Applications
Electrical and Electronic Engineering
Control and Optimization
Instrumentation

Access to Document

10.1109/ICCE-TW52618.2021.9603171

Cite this

Yeh, C. H., Chiu, D., Kang, L. W., Hsu, C. C., & Lo, C. (2021). Generative Adversarial Networks-based Face Hallucination with Identity-Preserving. In 2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021 (2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCE-TW52618.2021.9603171

Generative Adversarial Networks-based Face Hallucination with Identity-Preserving. / Yeh, Chia Hung; Chiu, Daniel; Kang, Li Wei et al.
2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021. Institute of Electrical and Electronics Engineers Inc., 2021. (2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Yeh, CH, Chiu, D, Kang, LW, Hsu, CC & Lo, C 2021, Generative Adversarial Networks-based Face Hallucination with Identity-Preserving. in 2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021. 2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021, Institute of Electrical and Electronics Engineers Inc., 8th IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021, Penghu, Taiwan, 2021/09/15. https://doi.org/10.1109/ICCE-TW52618.2021.9603171

Yeh CH, Chiu D, Kang LW, Hsu CC, Lo C. Generative Adversarial Networks-based Face Hallucination with Identity-Preserving. In 2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021. Institute of Electrical and Electronics Engineers Inc. 2021. (2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021). doi: 10.1109/ICCE-TW52618.2021.9603171

@inproceedings{01ddbb0bb99a4520b38f9812e5d3aa8e,

title = "Generative Adversarial Networks-based Face Hallucination with Identity-Preserving",

abstract = "This paper presents a novel generative adversarial networks-based face hallucination framework for producing high-resolution face images from very low-resolution (LR) ones. We propose a multi-scale generator architecture with multi-scale loss functions for different upscaling factors and a triplet-based identity preserving loss for extracting multi-scale identity-aware facial representations. Experimental results have verified that our method can well super-resolve very LR face images (e.g., 8×8) quantitatively and qualitatively.",

author = "Yeh, {Chia Hung} and Daniel Chiu and Kang, {Li Wei} and Hsu, {Chih Chung} and Chen Lo",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 8th IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021 ; Conference date: 15-09-2021 Through 17-09-2021",

year = "2021",

doi = "10.1109/ICCE-TW52618.2021.9603171",

language = "English",

series = "2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021",

}

TY - GEN

T1 - Generative Adversarial Networks-based Face Hallucination with Identity-Preserving

AU - Yeh, Chia Hung

AU - Chiu, Daniel

AU - Kang, Li Wei

AU - Hsu, Chih Chung

AU - Lo, Chen

PY - 2021

Y1 - 2021

N2 - This paper presents a novel generative adversarial networks-based face hallucination framework for producing high-resolution face images from very low-resolution (LR) ones. We propose a multi-scale generator architecture with multi-scale loss functions for different upscaling factors and a triplet-based identity preserving loss for extracting multi-scale identity-aware facial representations. Experimental results have verified that our method can well super-resolve very LR face images (e.g., 8×8) quantitatively and qualitatively.

AB - This paper presents a novel generative adversarial networks-based face hallucination framework for producing high-resolution face images from very low-resolution (LR) ones. We propose a multi-scale generator architecture with multi-scale loss functions for different upscaling factors and a triplet-based identity preserving loss for extracting multi-scale identity-aware facial representations. Experimental results have verified that our method can well super-resolve very LR face images (e.g., 8×8) quantitatively and qualitatively.

UR - http://www.scopus.com/inward/record.url?scp=85123054202&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85123054202&partnerID=8YFLogxK

U2 - 10.1109/ICCE-TW52618.2021.9603171

DO - 10.1109/ICCE-TW52618.2021.9603171

M3 - Conference contribution

AN - SCOPUS:85123054202

T3 - 2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021

BT - 2021 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 8th IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2021

Y2 - 15 September 2021 through 17 September 2021

ER -

Generative Adversarial Networks-based Face Hallucination with Identity-Preserving

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this