File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Learning Deep Representation for Face Alignment with Auxiliary Attributes

TitleLearning Deep Representation for Face Alignment with Auxiliary Attributes
Authors
Keywordsconvolutional network
deep learning
Face Alignment
face landmark detection
Issue Date2016
Citation
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, v. 38, n. 5, p. 918-930 How to Cite?
Abstract© 2015 IEEE. In this study, we show that landmark detection or face alignment task is not a single and independent problem. Instead, its robustness can be greatly improved with auxiliary information. Specifically, we jointly optimize landmark detection together with the recognition of heterogeneous but subtly correlated facial attributes, such as gender, expression, and appearance attributes. This is non-trivial since different attribute inference tasks have different learning difficulties and convergence rates. To address this problem, we formulate a novel tasks-constrained deep model, which not only learns the inter-task correlation but also employs dynamic task coefficients to facilitate the optimization convergence when learning multiple complex tasks. Extensive evaluations show that the proposed task-constrained learning (i) outperforms existing face alignment methods, especially in dealing with faces with severe occlusion and pose variation, and (ii) reduces model complexity drastically compared to the state-of-the-art methods based on cascaded deep model.
Persistent Identifierhttp://hdl.handle.net/10722/273556
ISSN
2023 Impact Factor: 20.8
2023 SCImago Journal Rankings: 6.158
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorZhang, Zhanpeng-
dc.contributor.authorLuo, Ping-
dc.contributor.authorLoy, Chen Change-
dc.contributor.authorTang, Xiaoou-
dc.date.accessioned2019-08-12T09:55:56Z-
dc.date.available2019-08-12T09:55:56Z-
dc.date.issued2016-
dc.identifier.citationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, v. 38, n. 5, p. 918-930-
dc.identifier.issn0162-8828-
dc.identifier.urihttp://hdl.handle.net/10722/273556-
dc.description.abstract© 2015 IEEE. In this study, we show that landmark detection or face alignment task is not a single and independent problem. Instead, its robustness can be greatly improved with auxiliary information. Specifically, we jointly optimize landmark detection together with the recognition of heterogeneous but subtly correlated facial attributes, such as gender, expression, and appearance attributes. This is non-trivial since different attribute inference tasks have different learning difficulties and convergence rates. To address this problem, we formulate a novel tasks-constrained deep model, which not only learns the inter-task correlation but also employs dynamic task coefficients to facilitate the optimization convergence when learning multiple complex tasks. Extensive evaluations show that the proposed task-constrained learning (i) outperforms existing face alignment methods, especially in dealing with faces with severe occlusion and pose variation, and (ii) reduces model complexity drastically compared to the state-of-the-art methods based on cascaded deep model.-
dc.languageeng-
dc.relation.ispartofIEEE Transactions on Pattern Analysis and Machine Intelligence-
dc.subjectconvolutional network-
dc.subjectdeep learning-
dc.subjectFace Alignment-
dc.subjectface landmark detection-
dc.titleLearning Deep Representation for Face Alignment with Auxiliary Attributes-
dc.typeArticle-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1109/TPAMI.2015.2469286-
dc.identifier.scopuseid_2-s2.0-84963829815-
dc.identifier.volume38-
dc.identifier.issue5-
dc.identifier.spage918-
dc.identifier.epage930-
dc.identifier.isiWOS:000374164700007-
dc.identifier.issnl0162-8828-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats