Generative Adversarial Network for Image Deblurring Using Content Constraint Loss

Ye Ji; Yaping Dai; Junjie Ma; Kaixin Zhao; Yanyan Cheng

doi:10.1109/CCDC49329.2020.9164664

Generative Adversarial Network for Image Deblurring Using Content Constraint Loss

Ye Ji, Yaping Dai, Junjie Ma, Kaixin Zhao, Yanyan Cheng

自动化学院

Beijing Institute of Technology

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

1 引用（Scopus）

摘要

The research of image deblurring plays an important role in the digital image processing. In order to reduce image blurring problems, a content constraint loss (CCL) function in the generative adversarial network (GAN) is proposed. The SSIM loss and the perceptual loss constitute the CCL function, which makes the trained generative model stable. The CCL function as the content constraint loss component and the adversarial loss component constitute the total loss. The total loss is optimized by the iterative training to further improve the stability of the network model, and the image blurring will be reduced. In the test experiment of the open source image dataset MNIST, CIFAR10/100 and CELEBA, the CCL function is used as the content constraint loss component of the generative adversarial network, the effect of image deblurring has obvious promotion in the structural similarity measure and visual appearance.

源语言	英语
主期刊名	Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020
出版商	Institute of Electrical and Electronics Engineers Inc.
页	1985-1990
页数	6
ISBN（电子版）	9781728158549
DOI	https://doi.org/10.1109/CCDC49329.2020.9164664
出版状态	已出版 - 8月 2020
活动	32nd Chinese Control and Decision Conference, CCDC 2020 - Hefei, 中国期限: 22 8月 2020 → 24 8月 2020

出版系列

姓名	Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020

会议

会议	32nd Chinese Control and Decision Conference, CCDC 2020
国家/地区	中国
市	Hefei
时期	22/08/20 → 24/08/20

访问文件

10.1109/CCDC49329.2020.9164664

其它文件与链接

链接到 Scopus 的出版物

引用此

Ji, Y., Dai, Y., Ma, J., Zhao, K., & Cheng, Y. (2020). Generative Adversarial Network for Image Deblurring Using Content Constraint Loss. 在 Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020 (页码 1985-1990). 文章 9164664 (Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CCDC49329.2020.9164664

@inproceedings{0c7d677d1e9a42fa958591965b805be2,

title = "Generative Adversarial Network for Image Deblurring Using Content Constraint Loss",

abstract = "The research of image deblurring plays an important role in the digital image processing. In order to reduce image blurring problems, a content constraint loss (CCL) function in the generative adversarial network (GAN) is proposed. The SSIM loss and the perceptual loss constitute the CCL function, which makes the trained generative model stable. The CCL function as the content constraint loss component and the adversarial loss component constitute the total loss. The total loss is optimized by the iterative training to further improve the stability of the network model, and the image blurring will be reduced. In the test experiment of the open source image dataset MNIST, CIFAR10/100 and CELEBA, the CCL function is used as the content constraint loss component of the generative adversarial network, the effect of image deblurring has obvious promotion in the structural similarity measure and visual appearance.",

keywords = "Adversarial loss, Content constraint loss, Generative adversarial network, Image deblurring",

author = "Ye Ji and Yaping Dai and Junjie Ma and Kaixin Zhao and Yanyan Cheng",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 32nd Chinese Control and Decision Conference, CCDC 2020 ; Conference date: 22-08-2020 Through 24-08-2020",

year = "2020",

month = aug,

doi = "10.1109/CCDC49329.2020.9164664",

language = "English",

series = "Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020",

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

pages = "1985--1990",

booktitle = "Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020",

address = "United States",

}

Ji, Y, Dai, Y, Ma, J, Zhao, K & Cheng, Y 2020, Generative Adversarial Network for Image Deblurring Using Content Constraint Loss. 在 Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020., 9164664, Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020, Institute of Electrical and Electronics Engineers Inc., 页码 1985-1990, 32nd Chinese Control and Decision Conference, CCDC 2020, Hefei, 中国, 22/08/20. https://doi.org/10.1109/CCDC49329.2020.9164664

Generative Adversarial Network for Image Deblurring Using Content Constraint Loss. / Ji, Ye; Dai, Yaping; Ma, Junjie 等.
Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020. Institute of Electrical and Electronics Engineers Inc., 2020. 页码 1985-1990 9164664 (Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020).

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

TY - GEN

T1 - Generative Adversarial Network for Image Deblurring Using Content Constraint Loss

AU - Ji, Ye

AU - Dai, Yaping

AU - Ma, Junjie

AU - Zhao, Kaixin

AU - Cheng, Yanyan

PY - 2020/8

Y1 - 2020/8

N2 - The research of image deblurring plays an important role in the digital image processing. In order to reduce image blurring problems, a content constraint loss (CCL) function in the generative adversarial network (GAN) is proposed. The SSIM loss and the perceptual loss constitute the CCL function, which makes the trained generative model stable. The CCL function as the content constraint loss component and the adversarial loss component constitute the total loss. The total loss is optimized by the iterative training to further improve the stability of the network model, and the image blurring will be reduced. In the test experiment of the open source image dataset MNIST, CIFAR10/100 and CELEBA, the CCL function is used as the content constraint loss component of the generative adversarial network, the effect of image deblurring has obvious promotion in the structural similarity measure and visual appearance.

AB - The research of image deblurring plays an important role in the digital image processing. In order to reduce image blurring problems, a content constraint loss (CCL) function in the generative adversarial network (GAN) is proposed. The SSIM loss and the perceptual loss constitute the CCL function, which makes the trained generative model stable. The CCL function as the content constraint loss component and the adversarial loss component constitute the total loss. The total loss is optimized by the iterative training to further improve the stability of the network model, and the image blurring will be reduced. In the test experiment of the open source image dataset MNIST, CIFAR10/100 and CELEBA, the CCL function is used as the content constraint loss component of the generative adversarial network, the effect of image deblurring has obvious promotion in the structural similarity measure and visual appearance.

KW - Adversarial loss

KW - Content constraint loss

KW - Generative adversarial network

KW - Image deblurring

UR - http://www.scopus.com/inward/record.url?scp=85091574133&partnerID=8YFLogxK

U2 - 10.1109/CCDC49329.2020.9164664

DO - 10.1109/CCDC49329.2020.9164664

M3 - Conference contribution

AN - SCOPUS:85091574133

T3 - Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020

SP - 1985

EP - 1990

BT - Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 32nd Chinese Control and Decision Conference, CCDC 2020

Y2 - 22 August 2020 through 24 August 2020

ER -

Ji Y, Dai Y, Ma J, Zhao K, Cheng Y. Generative Adversarial Network for Image Deblurring Using Content Constraint Loss. 在 Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020. Institute of Electrical and Electronics Engineers Inc. 2020. 页码 1985-1990. 9164664. (Proceedings of the 32nd Chinese Control and Decision Conference, CCDC 2020). doi: 10.1109/CCDC49329.2020.9164664

Generative Adversarial Network for Image Deblurring Using Content Constraint Loss

摘要

出版系列

会议

访问文件

其它文件与链接

指纹

引用此