CSpace  > 中国科学院计算技术研究所期刊论文  > 英文
Semantic invariant cross-domain image generation with generative adversarial networks
Mao, Xiaofeng1; Wang, Shuhui2; Zheng, Liying1; Huang, Qingming2,3
2018-06-07
发表期刊NEUROCOMPUTING
ISSN0925-2312
卷号293页码:55-63
摘要Recently, thanks to the state-of-the-art techniques in Generative Adversarial Networks, a lot of work achieves remarkable performance on learning the mapping between an input image and an output image without any paired relation. However, traditional methods on image-to-image translation merely consider the visual appearance properties, they fail to maintain the true semantics of an image during the transfer learning procedure from source to target domain. We propose a new approach that utilizes GAN to translate unpaired images between domains and remain high level semantic abstraction aligned. Our model controls the hierarchical semantics of images by processing semantic information on label level and spatial level respectively by constructing label and attention consistent losses. The experimental results on several benchmark datasets show that generated samples are both visually similar with target images and semantically consistent with their source counterparts. Furthermore, the experiment also suggests that our method can effectively improve the classification performance in unsupervised domain adaptation problem. (c) 2018 Elsevier B.V. All rights reserved.
关键词Generative adversarial networks Image-to-image translation Semantic invariance
DOI10.1016/j.neucom.2018.02.092
收录类别SCI
语种英语
资助项目National Natural Science Foundation of China[61771155] ; National Natural Science Foundation of China[61672497] ; National Natural Science Foundation of China[61332016] ; National Natural Science Foundation of China[61620106009] ; National Natural Science Foundation of China[61650202] ; National Natural Science Foundation of China[U1636214] ; National Basic Research Program of China (973 Program)[2015CB351802] ; Key Research Program of Frontier Sciences of CAS[QYZDJ-SSW-SYS013]
WOS研究方向Computer Science
WOS类目Computer Science, Artificial Intelligence
WOS记录号WOS:000429323200006
出版者ELSEVIER SCIENCE BV
引用统计
被引频次:35[WOS]   [WOS记录]     [WOS相关记录]
文献类型期刊论文
条目标识符http://119.78.100.204/handle/2XEOYT63/5745
专题中国科学院计算技术研究所期刊论文_英文
通讯作者Huang, Qingming
作者单位1.Harbin Engn Univ, Coll Comp Sci & Technol, Harbin, Heilongjiang, Peoples R China
2.Chinese Acad Sci, Inst Comp Technol, Key Lab Intell Info Proc, Beijing, Peoples R China
3.Univ Chinese Acad Sci, Sch Comp & Control Engn, Beijing, Peoples R China
推荐引用方式
GB/T 7714
Mao, Xiaofeng,Wang, Shuhui,Zheng, Liying,et al. Semantic invariant cross-domain image generation with generative adversarial networks[J]. NEUROCOMPUTING,2018,293:55-63.
APA Mao, Xiaofeng,Wang, Shuhui,Zheng, Liying,&Huang, Qingming.(2018).Semantic invariant cross-domain image generation with generative adversarial networks.NEUROCOMPUTING,293,55-63.
MLA Mao, Xiaofeng,et al."Semantic invariant cross-domain image generation with generative adversarial networks".NEUROCOMPUTING 293(2018):55-63.
条目包含的文件
条目无相关文件。
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
[Mao, Xiaofeng]的文章
[Wang, Shuhui]的文章
[Zheng, Liying]的文章
百度学术
百度学术中相似的文章
[Mao, Xiaofeng]的文章
[Wang, Shuhui]的文章
[Zheng, Liying]的文章
必应学术
必应学术中相似的文章
[Mao, Xiaofeng]的文章
[Wang, Shuhui]的文章
[Zheng, Liying]的文章
相关权益政策
暂无数据
收藏/分享
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。