Robust Kernelized Bayesian Matrix Factorization for Video Background/Foreground Separation

Xie, HB; Li, C; Xu, RYD; Mengersen, K

Robust Kernelized Bayesian Matrix Factorization for Video Background/Foreground Separation

Xie, HB Li, C

Xu, RYD

Mengersen, K

Permalink

Publication Type:: Chapter
Citation:: 2019, 11943 LNCS pp. 484 - 495
Issue Date:: 2019-01-01

Closed Access

	Filename	Description	Size
	Robust Kernelized Bayesian Matrix Factorization for Video BackgroundForeground Separation 1.pdf	Published version	734.49 kB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Xie, HB	en_US
dc.contributor.author	Li, C https://orcid.org/0000-0001-6600-2664	en_US
dc.contributor.author	Xu, RYD https://orcid.org/0000-0003-2080-4762	en_US
dc.contributor.author	Mengersen, K https://orcid.org/0000-0001-8625-9168	en_US
dc.date.issued	2019-01-01	en_US
dc.identifier.citation	2019, 11943 LNCS pp. 484 - 495	en_US
dc.identifier.isbn	9783030375980	en_US
dc.identifier.uri	http://hdl.handle.net/10453/138759
dc.description.abstract	© Springer Nature Switzerland AG 2019. Development of effective and efficient techniques for video analysis is an important research area in machine learning and computer vision. Matrix factorization (MF) is a powerful tool to perform such tasks. In this contribution, we present a hierarchical robust kernelized Bayesian matrix factorization (RKBMF) model to decompose a data set into low rank and sparse components. The RKBMF model automatically infers the parameters and latent variables including the reduced rank using variational Bayesian inference. Moreover, the model integrates the side information of similarity between frames to improve information extraction from the video. We employ RKBMF to extract background and foreground information from a traffic video. Experimental results demonstrate that RKBMF outperforms state-of-the-art approaches for background/foreground separation, particularly where the video is contaminated.	en_US
dc.relation.isbasedon	10.1007/978-3-030-37599-7_40	en_US
dc.subject.classification	Artificial Intelligence & Image Processing	en_US
dc.title	Robust Kernelized Bayesian Matrix Factorization for Video Background/Foreground Separation	en_US
dc.type	Chapter
utslib.citation.volume	11943 LNCS	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Students
utslib.copyright.status	closed_access
pubs.publication-status	Published	en_US
pubs.volume	11943 LNCS	en_US

Abstract:

© Springer Nature Switzerland AG 2019. Development of effective and efficient techniques for video analysis is an important research area in machine learning and computer vision. Matrix factorization (MF) is a powerful tool to perform such tasks. In this contribution, we present a hierarchical robust kernelized Bayesian matrix factorization (RKBMF) model to decompose a data set into low rank and sparse components. The RKBMF model automatically infers the parameters and latent variables including the reduced rank using variational Bayesian inference. Moreover, the model integrates the side information of similarity between frames to improve information extraction from the video. We employ RKBMF to extract background and foreground information from a traffic video. Experimental results demonstrate that RKBMF outperforms state-of-the-art approaches for background/foreground separation, particularly where the video is contaminated.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/138759