Show simple item record

dc.contributor.authorGao, Yang
dc.contributor.authorLi, Yi-Fan
dc.contributor.authorChandra, Swarup
dc.contributor.authorKhan, Latifur
dc.contributor.authorThuraisingham, Bhavani
dc.date.accessioned2020-03-11T22:19:54Z
dc.date.available2020-03-11T22:19:54Z
dc.date.issued2019-05
dc.identifier.isbn9781450366748
dc.identifier.urihttp://dx.doi.org/10.1145/3308558.3313503
dc.identifier.urihttps://hdl.handle.net/10735.1/7392
dc.description.abstractGood quality similarity metrics can significantly facilitate the performance of many large-scale, real-world applications. Existing studies have proposed various solutions to learn a Mahalanobis or bilinear metric in an online fashion by either restricting distances between similar (dissimilar) pairs to be smaller (larger) than a given lower (upper) bound or requiring similar instances to be separated from dissimilar instances with a given margin. However, these linear metrics learned by leveraging fixed bounds or margins may not perform well in real-world applications, especially when data distributions are complex. We aim to address the open challenge of “Online Adaptive Metric Learning” (OAML) for learning adaptive metric functions on-the-fly. Unlike traditional online metric learning methods, OAML is significantly more challenging since the learned metric could be non-linear and the model has to be self-adaptive as more instances are observed. In this paper, we present a new online metric learning framework that attempts to tackle the challenge by learning a ANN-based metric with adaptive model complexity from a stream of constraints. In particular, we propose a novel Adaptive-Bound Triplet Loss (ABTL) to effectively utilize the input constraints, and present a novel Adaptive Hedge Update (AHU) method for online updating the model parameters. We empirically validates the effectiveness and efficacy of our framework on various applications such as real-world image classification, facial verification, and image retrieval. © 2019 IW3C2 (International World Wide Web Conference Committee), published under Creative Commons CC-BY 4.0 License.
dc.description.sponsorshipNSF award numbers: DMS-1737978, DGE 17236021. OAC-1828467; ARO award number: W911-NF-18-1-0249
dc.language.isoen
dc.publisherAssociation For Computing Machinery, Inc
dc.relation.isPartOfThe Web Conference 2019 - Proceedings of the World Wide Web Conference
dc.rightsCC BY 4.0 (Attribution)
dc.rights©2019
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/
dc.subjectModeling, Adaptive
dc.subjectMachine learning
dc.subjectImage registration
dc.subjectDigital image correlation
dc.titleTowards Self-Adaptive Metric Learning on the Fly
dc.type.genrearticle
dc.description.departmentErik Jonsson School of Engineering and Computer Science
dc.identifier.bibliographicCitationGao, Y., Y. -F Li, S. Chandra, L. Khan, et al. 2019. "Towards self-adaptive metric learning on the fly." The Web Conference 2019 - Proceedings of the World Wide Web Conference: 503-513, doi: 10.1145/3308558.3313503
dc.contributor.utdAuthorGao, Yang
dc.contributor.utdAuthorLi, Yi-Fan
dc.contributor.utdAuthorChandra, Swarup
dc.contributor.utdAuthorKhan, Latifur
dc.contributor.utdAuthorThuraisingham, Bhavani M.
dc.contributor.VIAF51867299 (Thuraisingham, BM)


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

CC BY 4.0 (Attribution)
Except where otherwise noted, this item's license is described as CC BY 4.0 (Attribution)