DocumentCode :
2437722
Title :
A variational statistical framework for clustering human action videos
Author :
Fan, Wentao ; Bouguila, Nizar
Author_Institution :
Concordia Inst. for Inf. Syst. Eng., Concordia Univ., Montreal, QC, Canada
fYear :
2012
fDate :
23-25 May 2012
Firstpage :
1
Lastpage :
4
Abstract :
In this paper, we present an unsupervised learning method, based on the finite Dirichlet mixture model and the bag-of-visual words representation, for categorizing human action videos. The proposed Bayesian model is learned through a principled variational framework. A variational form of the Deviance Information Criterion (DIC) is incorporated within the proposed statistical framework for evaluating the correctness of the model complexity (i.e. number of mixture components). The effectiveness of the proposed model is illustrated through empirical results.
Keywords :
Bayes methods; statistical analysis; variational techniques; video signal processing; Bayesian model; bag-of-visual words representation; deviance information criterion; finite Dirichlet mixture model; human action videos; model complexity; principled variational framework; unsupervised learning; variational statistical framework; Accuracy; Approximation methods; Feature extraction; Humans; Vectors; Videos; Visualization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Image Analysis for Multimedia Interactive Services (WIAMIS), 2012 13th International Workshop on
Conference_Location :
Dublin
ISSN :
2158-5873
Print_ISBN :
978-1-4673-0791-8
Electronic_ISBN :
2158-5873
Type :
conf
DOI :
10.1109/WIAMIS.2012.6226748
Filename :
6226748
Link To Document :
بازگشت