A Novel Image Tag Completion Method Based on Convolutional Neural Transformation

Handle URI:
http://hdl.handle.net/10754/626775
Title:
A Novel Image Tag Completion Method Based on Convolutional Neural Transformation
Authors:
Geng, Yanyan; Zhang, Guohui; Li, Weizhi; Gu, Yi; Liang, Ru-Ze; Liang, Gaoyuan; Wang, Jingbin; Wu, Yanbin; Patil, Nitin; Wang, Jing-Yan
Abstract:
In the problems of image retrieval and annotation, complete textual tag lists of images play critical roles. However, in real-world applications, the image tags are usually incomplete, thus it is important to learn the complete tags for images. In this paper, we study the problem of image tag complete and proposed a novel method for this problem based on a popular image representation method, convolutional neural network (CNN). The method estimates the complete tags from the convolutional filtering outputs of images based on a linear predictor. The CNN parameters, linear predictor, and the complete tags are learned jointly by our method. We build a minimization problem to encourage the consistency between the complete tags and the available incomplete tags, reduce the estimation error, and reduce the model complexity. An iterative algorithm is developed to solve the minimization problem. Experiments over benchmark image data sets show its effectiveness.
KAUST Department:
Physical Sciences and Engineering (PSE) Division; Materials Science and Engineering Program
Citation:
Geng Y, Zhang G, Li W, Gu Y, Liang R-Z, et al. (2017) A Novel Image Tag Completion Method Based on Convolutional Neural Transformation. Lecture Notes in Computer Science: 539–546. Available: http://dx.doi.org/10.1007/978-3-319-68612-7_61.
Publisher:
Springer International Publishing
Journal:
Artificial Neural Networks and Machine Learning – ICANN 2017
Conference/Event name:
26th International Conference on Artificial Neural Networks, ICANN 2017
Issue Date:
24-Oct-2017
DOI:
10.1007/978-3-319-68612-7_61
Type:
Conference Paper
ISSN:
0302-9743; 1611-3349
Additional Links:
https://link.springer.com/chapter/10.1007%2F978-3-319-68612-7_61
Appears in Collections:
Conference Papers; Physical Sciences and Engineering (PSE) Division; Materials Science and Engineering Program

Full metadata record

DC FieldValue Language
dc.contributor.authorGeng, Yanyanen
dc.contributor.authorZhang, Guohuien
dc.contributor.authorLi, Weizhien
dc.contributor.authorGu, Yien
dc.contributor.authorLiang, Ru-Zeen
dc.contributor.authorLiang, Gaoyuanen
dc.contributor.authorWang, Jingbinen
dc.contributor.authorWu, Yanbinen
dc.contributor.authorPatil, Nitinen
dc.contributor.authorWang, Jing-Yanen
dc.date.accessioned2018-01-15T06:35:08Z-
dc.date.available2018-01-15T06:35:08Z-
dc.date.issued2017-10-24en
dc.identifier.citationGeng Y, Zhang G, Li W, Gu Y, Liang R-Z, et al. (2017) A Novel Image Tag Completion Method Based on Convolutional Neural Transformation. Lecture Notes in Computer Science: 539–546. Available: http://dx.doi.org/10.1007/978-3-319-68612-7_61.en
dc.identifier.issn0302-9743en
dc.identifier.issn1611-3349en
dc.identifier.doi10.1007/978-3-319-68612-7_61en
dc.identifier.urihttp://hdl.handle.net/10754/626775-
dc.description.abstractIn the problems of image retrieval and annotation, complete textual tag lists of images play critical roles. However, in real-world applications, the image tags are usually incomplete, thus it is important to learn the complete tags for images. In this paper, we study the problem of image tag complete and proposed a novel method for this problem based on a popular image representation method, convolutional neural network (CNN). The method estimates the complete tags from the convolutional filtering outputs of images based on a linear predictor. The CNN parameters, linear predictor, and the complete tags are learned jointly by our method. We build a minimization problem to encourage the consistency between the complete tags and the available incomplete tags, reduce the estimation error, and reduce the model complexity. An iterative algorithm is developed to solve the minimization problem. Experiments over benchmark image data sets show its effectiveness.en
dc.publisherSpringer International Publishingen
dc.relation.urlhttps://link.springer.com/chapter/10.1007%2F978-3-319-68612-7_61en
dc.subjectConvolutional neural filteringen
dc.subjectImage representationen
dc.subjectTag completionen
dc.subjectImage retrievalen
dc.subjectImage annotationen
dc.titleA Novel Image Tag Completion Method Based on Convolutional Neural Transformationen
dc.typeConference Paperen
dc.contributor.departmentPhysical Sciences and Engineering (PSE) Divisionen
dc.contributor.departmentMaterials Science and Engineering Programen
dc.identifier.journalArtificial Neural Networks and Machine Learning – ICANN 2017en
dc.conference.date2017-09-11 to 2017-09-14en
dc.conference.name26th International Conference on Artificial Neural Networks, ICANN 2017en
dc.conference.locationAlghero, ITAen
dc.contributor.institutionProvincial Key Laboratory for Computer Information Processing Technology, Soochow University, Suzhou, Chinaen
dc.contributor.institutionHuawei Technologies Co., Ltd., Shanghai, Chinaen
dc.contributor.institutionSuning Commerce R&D Center USA, Inc., Palo Alto, USAen
dc.contributor.institutionAnalytics and Research, Travelers, Hartford, USAen
dc.contributor.institutionJiangsu University of Technology, Jiangsu, Chinaen
dc.contributor.institutionInformation Technology Service Center, Intermediate People’s Court of Linyi City, Linyi, Chinaen
dc.contributor.institutionHebei University of Economics and Business, Shijiazhuang, Chinaen
dc.contributor.institutionSavitribai Phule Pune University, Pune, Indiaen
dc.contributor.institutionJiangsu Key Laboratory of Big Data Analysis Technology/B-DAT, Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing, Chinaen
kaust.authorLiang, Ru-Zeen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.