DehazeGAN: Underwater Haze Image Restoration using Unpaired Image-to-image Translation

Younggun Cho; Ramavtar Malav; Gaurav Pandey; Ayoung Kim

doi:10.1016/j.ifacol.2019.12.287

DehazeGAN: Underwater Haze Image Restoration using Unpaired Image-to-image Translation

Younggun Cho, Ramavtar Malav, Gaurav Pandey, Ayoung Kim

Research output: Contribution to journal › Conference article › peer-review

8 Scopus citations

Abstract

In this paper, we propose a Generative Adversarial Networks (GAN)-based image restoration method. Our method adopts an unpaired image-to-image translation network to learn the characteristics of underwater haze images. To enhance restoration, we propose multiple cyclic consistency losses that capture the detail of images and suppress distortion image translation. To prepare unpaired images of clean and degraded scenes, we collected images from Flickr and filter out false images using image characteristics. The proposed network is tested on public underwater images and shows promising results under severe image distortion.

Original language	English
Pages (from-to)	82-85
Number of pages	4
Journal	IFAC-PapersOnLine
Volume	52
Issue number	21
DOIs	https://doi.org/10.1016/j.ifacol.2019.12.287
State	Published - 2019
Externally published	Yes
Event	12th IFAC Conference on Control Applications in Marine Systems, Robotics, and Vehicles CAMS 2019 - Daejeon, Korea, Republic of Duration: 18 Sep 2019 → 20 Sep 2019

Bibliographical note

Publisher Copyright:
© 2019. The Authors. Published by Elsevier Ltd. All rights reserved.

Access to Document

10.1016/j.ifacol.2019.12.287

Cite this

@article{beafd0e618fb468e8e9fb86318fb0a03,

title = "DehazeGAN: Underwater Haze Image Restoration using Unpaired Image-to-image Translation",

abstract = "In this paper, we propose a Generative Adversarial Networks (GAN)-based image restoration method. Our method adopts an unpaired image-to-image translation network to learn the characteristics of underwater haze images. To enhance restoration, we propose multiple cyclic consistency losses that capture the detail of images and suppress distortion image translation. To prepare unpaired images of clean and degraded scenes, we collected images from Flickr and filter out false images using image characteristics. The proposed network is tested on public underwater images and shows promising results under severe image distortion.",

author = "Younggun Cho and Ramavtar Malav and Gaurav Pandey and Ayoung Kim",

note = "Publisher Copyright: {\textcopyright} 2019. The Authors. Published by Elsevier Ltd. All rights reserved.; 12th IFAC Conference on Control Applications in Marine Systems, Robotics, and Vehicles CAMS 2019 ; Conference date: 18-09-2019 Through 20-09-2019",

year = "2019",

doi = "10.1016/j.ifacol.2019.12.287",

language = "English",

volume = "52",

pages = "82--85",

journal = "IFAC-PapersOnLine",

issn = "2405-8963",

publisher = "IFAC Secretariat",

number = "21",

}

TY - JOUR

T1 - DehazeGAN

T2 - 12th IFAC Conference on Control Applications in Marine Systems, Robotics, and Vehicles CAMS 2019

AU - Cho, Younggun

AU - Malav, Ramavtar

AU - Pandey, Gaurav

AU - Kim, Ayoung

PY - 2019

Y1 - 2019

N2 - In this paper, we propose a Generative Adversarial Networks (GAN)-based image restoration method. Our method adopts an unpaired image-to-image translation network to learn the characteristics of underwater haze images. To enhance restoration, we propose multiple cyclic consistency losses that capture the detail of images and suppress distortion image translation. To prepare unpaired images of clean and degraded scenes, we collected images from Flickr and filter out false images using image characteristics. The proposed network is tested on public underwater images and shows promising results under severe image distortion.

AB - In this paper, we propose a Generative Adversarial Networks (GAN)-based image restoration method. Our method adopts an unpaired image-to-image translation network to learn the characteristics of underwater haze images. To enhance restoration, we propose multiple cyclic consistency losses that capture the detail of images and suppress distortion image translation. To prepare unpaired images of clean and degraded scenes, we collected images from Flickr and filter out false images using image characteristics. The proposed network is tested on public underwater images and shows promising results under severe image distortion.

UR - http://www.scopus.com/inward/record.url?scp=85079638919&partnerID=8YFLogxK

U2 - 10.1016/j.ifacol.2019.12.287

DO - 10.1016/j.ifacol.2019.12.287

M3 - Conference article

AN - SCOPUS:85079638919

SN - 2405-8963

VL - 52

SP - 82

EP - 85

JO - IFAC-PapersOnLine

JF - IFAC-PapersOnLine

IS - 21

Y2 - 18 September 2019 through 20 September 2019

ER -

DehazeGAN: Underwater Haze Image Restoration using Unpaired Image-to-image Translation

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this