File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Social image annotation via cross-domain subspace learning

TitleSocial image annotation via cross-domain subspace learning
Authors
KeywordsCross-Domain Learning
Social Image Annotation
Subspace Learning
Issue Date2012
PublisherSpringer New York LLC. The Journal's web site is located at http://springerlink.metapress.com/openurl.asp?genre=journal&issn=1380-7501
Citation
Multimedia Tools And Applications, 2012, v. 56 n. 1, p. 91-108 How to Cite?
AbstractIn recent years, cross-domain learning algorithms have attracted much attention to solve labeled data insufficient problem. However, these cross-domain learning algorithms cannot be applied for subspace learning, which plays a key role in multimedia processing. This paper envisions the cross-domain discriminative subspace learning and provides an effective solution to cross-domain subspace learning. In particular, we propose the cross-domain discriminative locally linear embedding or CDLLE for short. CDLLE connects the training and the testing samples by minimizing the quadratic distance between the distribution of the training samples and that of the testing samples. Therefore, a common subspace for data representation can be preserved. We basically expect the discriminative information to separate the concepts in the training set can be shared to separate the concepts in the testing set as well and thus we have a chance to address above cross-domain problem duly. The margin maximization is duly adopted in CDLLE so the discriminative information for separating different classes can be well preserved. Finally, CDLLE encodes the local geometry of each training samples through a series of linear coefficients which can reconstruct a given sample by its intra-class neighbour samples and thus can locally preserve the intra-class local geometry. Experimental evidence on NUS-WIDE, a popular social image database collected from Flickr, and MSRA-MM, a popular real-world web image annotation database collected from the Internet by using Microsoft Live Search, demonstrates the effectiveness of CDLLE for real-world cross-domain applications. © 2010 Springer Science+Business Media, LLC.
Persistent Identifierhttp://hdl.handle.net/10722/152495
ISSN
2023 Impact Factor: 3.0
2023 SCImago Journal Rankings: 0.801
ISI Accession Number ID
References

 

DC FieldValueLanguage
dc.contributor.authorSi, Sen_US
dc.contributor.authorTao, Den_US
dc.contributor.authorWang, Men_US
dc.contributor.authorChan, KPen_US
dc.date.accessioned2012-06-26T06:39:40Z-
dc.date.available2012-06-26T06:39:40Z-
dc.date.issued2012en_US
dc.identifier.citationMultimedia Tools And Applications, 2012, v. 56 n. 1, p. 91-108en_US
dc.identifier.issn1380-7501en_US
dc.identifier.urihttp://hdl.handle.net/10722/152495-
dc.description.abstractIn recent years, cross-domain learning algorithms have attracted much attention to solve labeled data insufficient problem. However, these cross-domain learning algorithms cannot be applied for subspace learning, which plays a key role in multimedia processing. This paper envisions the cross-domain discriminative subspace learning and provides an effective solution to cross-domain subspace learning. In particular, we propose the cross-domain discriminative locally linear embedding or CDLLE for short. CDLLE connects the training and the testing samples by minimizing the quadratic distance between the distribution of the training samples and that of the testing samples. Therefore, a common subspace for data representation can be preserved. We basically expect the discriminative information to separate the concepts in the training set can be shared to separate the concepts in the testing set as well and thus we have a chance to address above cross-domain problem duly. The margin maximization is duly adopted in CDLLE so the discriminative information for separating different classes can be well preserved. Finally, CDLLE encodes the local geometry of each training samples through a series of linear coefficients which can reconstruct a given sample by its intra-class neighbour samples and thus can locally preserve the intra-class local geometry. Experimental evidence on NUS-WIDE, a popular social image database collected from Flickr, and MSRA-MM, a popular real-world web image annotation database collected from the Internet by using Microsoft Live Search, demonstrates the effectiveness of CDLLE for real-world cross-domain applications. © 2010 Springer Science+Business Media, LLC.en_US
dc.languageengen_US
dc.publisherSpringer New York LLC. The Journal's web site is located at http://springerlink.metapress.com/openurl.asp?genre=journal&issn=1380-7501en_US
dc.relation.ispartofMultimedia Tools and Applicationsen_US
dc.subjectCross-Domain Learningen_US
dc.subjectSocial Image Annotationen_US
dc.subjectSubspace Learningen_US
dc.titleSocial image annotation via cross-domain subspace learningen_US
dc.typeArticleen_US
dc.identifier.emailChan, KP:kpchan@cs.hku.hken_US
dc.identifier.authorityChan, KP=rp00092en_US
dc.description.naturelink_to_subscribed_fulltexten_US
dc.identifier.doi10.1007/s11042-010-0567-2en_US
dc.identifier.scopuseid_2-s2.0-84857361283en_US
dc.relation.referenceshttp://www.scopus.com/mlt/select.url?eid=2-s2.0-84857361283&selection=ref&src=s&origin=recordpageen_US
dc.identifier.volume56en_US
dc.identifier.issue1en_US
dc.identifier.spage91en_US
dc.identifier.epage108en_US
dc.identifier.isiWOS:000299127500005-
dc.publisher.placeUnited Statesen_US
dc.identifier.scopusauthoridSi, S=35422764200en_US
dc.identifier.scopusauthoridTao, D=7102600334en_US
dc.identifier.scopusauthoridWang, M=7406689641en_US
dc.identifier.scopusauthoridChan, KP=7406032820en_US
dc.identifier.citeulike7775436-
dc.identifier.issnl1380-7501-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats