Local Smoothing Constraint in Convolutional Neural Network for Image Denoising

Yonghong Guo; Feng Jiang; Dongyang Zhao; Zhaojing Wen; Shaohui Liu

doi:10.1007/978-3-030-24274-9_36

Local Smoothing Constraint in Convolutional Neural Network for Image Denoising

Yonghong Guo^*
, Feng Jiang
, Dongyang Zhao
, Zhaojing Wen
, Shaohui Liu

^*Corresponding author for this work

Beijing Institute of Computer Application
School of Computer Science and Technology, Harbin Institute of Technology

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

Abstract

In this paper, we demonstrate that not only natural images but also their intermediate responses of convolutional neural networks (CNNs) have local smoothing priors. To imposing the local smoothing constraint, we design a local smoothing layer, which is able to suppress noises in a local receptive field. Further, we arrange the local smoothing layer in the early layers of CNNs to effectively capture context information, which is helpful to recovery image details. Experimental results validate that the proposed denoising method outperforms several state-of-the-art methods.

Original language	English
Title of host publication	Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings
Editors	Xingming Sun, Zhaoqing Pan, Elisa Bertino
Publisher	Springer Verlag
Pages	402-410
Number of pages	9
ISBN (Print)	9783030242732
DOIs	https://doi.org/10.1007/978-3-030-24274-9_36
State	Published - 2019
Externally published	Yes
Event	5th International Conference on Artificial Intelligence and Security, ICAIS 2019 - New York city, United States Duration: 26 Jul 2019 → 28 Jul 2019

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11632 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	5th International Conference on Artificial Intelligence and Security, ICAIS 2019
Country/Territory	United States
City	New York city
Period	26/07/19 → 28/07/19

Keywords

Convolutional neural network
Denoise
Smoothing priors

Access to Document

10.1007/978-3-030-24274-9_36

Cite this

Guo, Y., Jiang, F., Zhao, D., Wen, Z., & Liu, S. (2019). Local Smoothing Constraint in Convolutional Neural Network for Image Denoising. In X. Sun, Z. Pan, & E. Bertino (Eds.), Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings (pp. 402-410). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11632 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-24274-9_36

Guo, Yonghong ; Jiang, Feng ; Zhao, Dongyang et al. / Local Smoothing Constraint in Convolutional Neural Network for Image Denoising. Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings. editor / Xingming Sun ; Zhaoqing Pan ; Elisa Bertino. Springer Verlag, 2019. pp. 402-410 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3962f0e66a0a4f5ba803d6312c10ded1,

title = "Local Smoothing Constraint in Convolutional Neural Network for Image Denoising",

abstract = "In this paper, we demonstrate that not only natural images but also their intermediate responses of convolutional neural networks (CNNs) have local smoothing priors. To imposing the local smoothing constraint, we design a local smoothing layer, which is able to suppress noises in a local receptive field. Further, we arrange the local smoothing layer in the early layers of CNNs to effectively capture context information, which is helpful to recovery image details. Experimental results validate that the proposed denoising method outperforms several state-of-the-art methods.",

keywords = "Convolutional neural network, Denoise, Smoothing priors",

author = "Yonghong Guo and Feng Jiang and Dongyang Zhao and Zhaojing Wen and Shaohui Liu",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 5th International Conference on Artificial Intelligence and Security, ICAIS 2019 ; Conference date: 26-07-2019 Through 28-07-2019",

year = "2019",

doi = "10.1007/978-3-030-24274-9\_36",

language = "英语",

isbn = "9783030242732",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "402--410",

editor = "Xingming Sun and Zhaoqing Pan and Elisa Bertino",

booktitle = "Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings",

address = "德国",

}

Guo, Y, Jiang, F, Zhao, D, Wen, Z & Liu, S 2019, Local Smoothing Constraint in Convolutional Neural Network for Image Denoising. in X Sun, Z Pan & E Bertino (eds), Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11632 LNCS, Springer Verlag, pp. 402-410, 5th International Conference on Artificial Intelligence and Security, ICAIS 2019, New York city, United States, 26/07/19. https://doi.org/10.1007/978-3-030-24274-9_36

Local Smoothing Constraint in Convolutional Neural Network for Image Denoising. / Guo, Yonghong; Jiang, Feng; Zhao, Dongyang et al.
Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings. ed. / Xingming Sun; Zhaoqing Pan; Elisa Bertino. Springer Verlag, 2019. p. 402-410 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11632 LNCS).

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

TY - GEN

T1 - Local Smoothing Constraint in Convolutional Neural Network for Image Denoising

AU - Guo, Yonghong

AU - Jiang, Feng

AU - Zhao, Dongyang

AU - Wen, Zhaojing

AU - Liu, Shaohui

PY - 2019

Y1 - 2019

N2 - In this paper, we demonstrate that not only natural images but also their intermediate responses of convolutional neural networks (CNNs) have local smoothing priors. To imposing the local smoothing constraint, we design a local smoothing layer, which is able to suppress noises in a local receptive field. Further, we arrange the local smoothing layer in the early layers of CNNs to effectively capture context information, which is helpful to recovery image details. Experimental results validate that the proposed denoising method outperforms several state-of-the-art methods.

AB - In this paper, we demonstrate that not only natural images but also their intermediate responses of convolutional neural networks (CNNs) have local smoothing priors. To imposing the local smoothing constraint, we design a local smoothing layer, which is able to suppress noises in a local receptive field. Further, we arrange the local smoothing layer in the early layers of CNNs to effectively capture context information, which is helpful to recovery image details. Experimental results validate that the proposed denoising method outperforms several state-of-the-art methods.

KW - Convolutional neural network

KW - Denoise

KW - Smoothing priors

UR - https://www.scopus.com/pages/publications/85073902473

U2 - 10.1007/978-3-030-24274-9_36

DO - 10.1007/978-3-030-24274-9_36

M3 - 会议稿件

AN - SCOPUS:85073902473

SN - 9783030242732

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 402

EP - 410

BT - Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings

A2 - Sun, Xingming

A2 - Pan, Zhaoqing

A2 - Bertino, Elisa

PB - Springer Verlag

T2 - 5th International Conference on Artificial Intelligence and Security, ICAIS 2019

Y2 - 26 July 2019 through 28 July 2019

ER -

Guo Y, Jiang F, Zhao D, Wen Z, Liu S. Local Smoothing Constraint in Convolutional Neural Network for Image Denoising. In Sun X, Pan Z, Bertino E, editors, Artificial Intelligence and Security - 5th International Conference, ICAIS 2019, Proceedings. Springer Verlag. 2019. p. 402-410. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-24274-9_36

Local Smoothing Constraint in Convolutional Neural Network for Image Denoising

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this