File(s) under permanent embargo
Improved subspace clustering via exploitation of spatial constraints
conference contribution
posted on 2012-01-01, 00:00 authored by D S Pham, Budhaditya Saha, Quoc-Dinh Phung, Svetha VenkateshSvetha VenkateshWe present a novel approach to improving subspace clustering by exploiting the spatial constraints. The new method encourages the sparse solution to be consistent with the spatial geometry of the tracked points, by embedding weights into the sparse formulation. By doing so, we are able to correct sparse representations in a principled manner without introducing much additional computational cost. We discuss alternative ways to treat the missing and corrupted data using the latest theory in robust lasso regression and suggest numerical algorithms so solve the proposed formulation. The experiments on the benchmark Johns Hopkins 155 dataset demonstrate that exploiting spatial constraints significantly improves motion segmentation.
History
Event
Computer Vision and Pattern Recognition. Conference (2012 : Providence, R. I.)Pagination
550 - 557Publisher
IEEELocation
Providence, R. I.Place of publication
Piscataway, N. J.Start date
2012-06-16End date
2012-06-21ISSN
1063-6919ISBN-13
9781467312264ISBN-10
1467312266Language
engPublication classification
E1 Full written paper - refereedTitle of proceedings
CVPR 2012 : Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern RecognitionUsage metrics
Categories
No categories selectedKeywords
computational costscorrupted datadata setsHopkinsmotion segmentationnumerical algorithmssparse representationsparse solutionsspatial constraintsspatial geometrysubspace clusteringScience & TechnologyTechnologyComputer Science, Artificial IntelligenceComputer Science, Interdisciplinary ApplicationsEngineering, Electrical & ElectronicComputer ScienceEngineeringUNIONRECOVERYSIGNALSMODEL
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC