Back to Search Start Over

Model Selection for Unsupervised Learning of Visual Context.

Authors :
Xiang, Tao
Gong, Shaogang
Source :
International Journal of Computer Vision. Aug2006, Vol. 69 Issue 2, p181-201. 21p. 2 Color Photographs, 1 Black and White Photograph, 6 Charts, 6 Graphs.
Publication Year :
2006

Abstract

This study addresses the problem of choosing the most suitable probabilistic model selection criterion for unsupervised learning of visual context of a dynamic scene using mixture models. A rectified Bayesian Information Criterion (BICr) and a Completed Likelihood Akaike’s Information Criterion (CL-AIC) are formulated to estimate the optimal model order (complexity) for a given visual scene. Both criteria are designed to overcome poor model selection by existing popular criteria when the data sample size varies from small to large and the true mixture distribution kernel functions differ from the assumed ones. Extensive experiments on learning visual context for dynamic scene modelling are carried out to demonstrate the effectiveness of BICr and CL-AIC, compared to that of existing popular model selection criteria including BIC, AIC and Integrated Completed Likelihood (ICL). Our study suggests that for learning visual context using a mixture model, BICr is the most appropriate criterion given sparse data, while CL-AIC should be chosen given moderate or large data sample sizes. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
09205691
Volume :
69
Issue :
2
Database :
Academic Search Index
Journal :
International Journal of Computer Vision
Publication Type :
Academic Journal
Accession number :
20907676
Full Text :
https://doi.org/10.1007/s11263-005-5024-8