Towards Real-Time Advancement of Underwater Visual Quality With GAN
Chen, Xingyu1,2; Yu, Junzhi1,2,3; Kong, Shihan1,2; Wu, Zhengxing1,2; Fang, Xi4; Wen, Li4
发表期刊IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
ISSN0278-0046
2019-12-01
卷号66期号:12页码:9350-9359
通讯作者Yu, Junzhi(junzhi.yu@ia.ac.cn)
摘要

Low visual quality has prevented underwater robotic vision from a wide range of applications. Although several algorithms have been developed, real time and adaptive methods are deficient for real-world tasks. In this paper, we address this difficulty based on generative adversarial networks (GAN), and propose a GAN-based restoration scheme (GAN-RS). In particular, we develop a multibranch discriminator including an adversarial branch and a critic branch for the purpose of simultaneously preserving image content and removing underwater noise. In addition to adversarial learning, a novel dark channel prior loss also promotes the generator to produce realistic vision. More specifically, an underwater index is investigated to describe underwater properties, and a loss function based on the underwater index is designed to train the critic branch for underwater noise suppression. Through extensive comparisons on visual quality and feature restoration, we confirm the superiority of the proposed approach. Consequently, the GAN-RS can adaptively improve underwater visual quality in real time and induce an overall superior restoration performance. Finally, a real-world experiment is conducted on the seabed for grasping marine products, and the results are quite promising. The source code is publicly available(1).

关键词Generative adversarial networks (GAN) image restoration machine learning underwater vision
DOI10.1109/TIE.2019.2893840
关键词[WOS]IMAGE-ENHANCEMENT
收录类别SCI
语种英语
资助项目National Natural Science Foundation of China[61633004] ; National Natural Science Foundation of China[61725305] ; National Natural Science Foundation of China[61633017] ; National Natural Science Foundation of China[61603388] ; Beijing Natural Science Foundation[4161002] ; Beijing Natural Science Foundation[4161002] ; National Natural Science Foundation of China[61603388] ; National Natural Science Foundation of China[61633017] ; National Natural Science Foundation of China[61725305] ; National Natural Science Foundation of China[61633004]
项目资助者National Natural Science Foundation of China ; Beijing Natural Science Foundation
WOS研究方向Automation & Control Systems ; Engineering ; Instruments & Instrumentation
WOS类目Automation & Control Systems ; Engineering, Electrical & Electronic ; Instruments & Instrumentation
WOS记录号WOS:000480309400023
出版者IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
被引频次:82[WOS]   [WOS记录]     [WOS相关记录]
文献类型期刊论文
条目标识符http://ir.ia.ac.cn/handle/173211/27563
专题复杂系统管理与控制国家重点实验室
通讯作者Yu, Junzhi
作者单位1.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China
2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China
3.Peking Univ, Beijing Innovat Ctr Engn Sci & Adv Technol, Beijing 100871, Peoples R China
4.Beihang Univ, Sch Mech Engn & Automat, Beijing 100191, Peoples R China
第一作者单位中国科学院自动化研究所
通讯作者单位中国科学院自动化研究所
推荐引用方式
GB/T 7714
Chen, Xingyu,Yu, Junzhi,Kong, Shihan,et al. Towards Real-Time Advancement of Underwater Visual Quality With GAN[J]. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,2019,66(12):9350-9359.
APA Chen, Xingyu,Yu, Junzhi,Kong, Shihan,Wu, Zhengxing,Fang, Xi,&Wen, Li.(2019).Towards Real-Time Advancement of Underwater Visual Quality With GAN.IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,66(12),9350-9359.
MLA Chen, Xingyu,et al."Towards Real-Time Advancement of Underwater Visual Quality With GAN".IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 66.12(2019):9350-9359.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
GAN-RS_TIE.pdf(4984KB)期刊论文作者接受稿开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Chen, Xingyu]的文章
[Yu, Junzhi]的文章
[Kong, Shihan]的文章
百度学术
百度学术中相似的文章
[Chen, Xingyu]的文章
[Yu, Junzhi]的文章
[Kong, Shihan]的文章
必应学术
必应学术中相似的文章
[Chen, Xingyu]的文章
[Yu, Junzhi]的文章
[Kong, Shihan]的文章
相关权益政策
暂无数据
收藏/分享
文件名: GAN-RS_TIE.pdf
格式: Adobe PDF
此文件暂不支持浏览
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。