Learning Robust and Discriminative Subspace With Low-Rank Constraints

dc.contributor.authorSheng Li
dc.contributor.authorYun Fu
dc.date.accessioned2017-03-27T23:29:51Z
dc.date.available2017-03-27T23:29:51Z
dc.date.issued2016
dc.descriptionIEEE Transactions on Neural Networks and Learning Systemsen_US
dc.descriptionThe article of record as published may be found at http://dx.doi.org/10.1109/tnnls.2015.2464090en_US
dc.description.abstractIn this paper, we aim at learning robust and discriminative subspaces from noisy data. Subspace learning is widely used in extracting discriminative features for classifica- tion. However, when data are contaminated with severe noise, the performance of most existing subspace learning methods would be limited. Recent advances in low-rank modeling provide effective solutions for removing noise or outliers contained in sample sets, which motivates us to take advantage of low-rank constraints in order to exploit robust and discriminative subspace for classification. In particular, we present a discriminative subspace learning method called the supervised regularization- based robust subspace (SRRS) approach, by incorporating the low-rank constraint. SRRS seeks low-rank representations from the noisy data, and learns a discriminative subspace from the recovered clean data jointly. A supervised regularization function is designed to make use of the class label information, and therefore to enhance the discriminability of subspace. Our approach is formulated as a constrained rank-minimization problem. We design an inexact augmented Lagrange multiplier optimization algorithm to solve it. Unlike the existing sparse representation and low-rank learning methods, our approach learns a low-dimensional subspace from recovered data, and explicitly incorporates the supervised information. Our approach and some baselines are evaluated on the COIL-100, ALOI, Extended YaleB, FERET, AR, and KinFace databases. The exper- imental results demonstrate the effectiveness of our approach, especially when the data contain considerable noise or variations.en_US
dc.description.sponsorshipFunded by Naval Postgraduate Schoolen_US
dc.description.sponsorshipNational Science Foundation Computer and Network Systemsen_US
dc.description.sponsorshipONR Young Investigatoren_US
dc.description.sponsorshipOffice of Naval Researchen_US
dc.description.sponsorshipU.S. Army Research Office Young Investigatoren_US
dc.identifier.urihttps://hdl.handle.net/10945/52406
dc.relation.ispartofseriesFunded by Naval Postgraduate School
dc.rightsCopyright is reserved by the copyright owner.en_US
dc.titleLearning Robust and Discriminative Subspace With Low-Rank Constraintsen_US
dc.typeArticleen_US
dspace.entity.typePublication
relation.isSeriesOfPublicationaf4f559a-d2da-4fa4-89d1-78532cf33472
relation.isSeriesOfPublication.latestForDiscoveryaf4f559a-d2da-4fa4-89d1-78532cf33472
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Learning_Robust_and_Discriminative.pdf
Size:
3.02 MB
Format:
Adobe Portable Document Format
Collections