Paper Info
Title
Clustering Dynamic Textures with the Hierarchical EM Algorithm for Modeling Video
Abstract
Dynamic texture (DT) is a probabilistic generative model, defined over space and time, that represents a video as the output of a linear dynamical system (LDS). The DT model has been applied to a wide variety of computer vision problems, such as motion segmentation, motion classification, and video registration. In this paper, we derive a new algorithm for clustering DT models that is based on the hierarchical EM algorithm. The proposed clustering algorithm is capable of both clustering DTs and learning novel DT cluster centers that are representative of the cluster members in a manner that is consistent with the underlying generative probabilistic model of the DT. We also derive an efficient recursive algorithm for sensitivity analysis of the discrete-time Kalman smoothing filter, which is used as the basis for computing expectations in the E-step of the HEM algorithm. Finally, we demonstrate the efficacy of the clustering algorithm on several applications in motion analysis, including hierarchical motion clustering, semantic motion annotation, and learning bag-of-systems (BoS) codebooks for dynamic texture recognition.
Year
DOI
Venue
2013
10.1109/TPAMI.2012.236
Pattern Analysis and Machine Intelligence, IEEE Transactions
Keywords
Field
DocType
Kalman filters,computer vision,discrete time systems,expectation-maximisation algorithm,image texture,learning (artificial intelligence),pattern clustering,probability,sensitivity analysis,smoothing methods,video signal processing,BoS codebook learning,DT cluster center learning,DT model,HEM algorithm,LDS,bag-of-system codebook learning,cluster members,computer vision problems,discrete-time Kalman smoothing filter,dynamic texture clustering algorithm,dynamic texture recognition,hierarchical EM algorithm,hierarchical motion clustering,linear dynamical system,motion classification,motion segmentation,probabilistic generative model,recursive algorithm,semantic motion annotation,sensitivity analysis,video modeling,video registration,Dynamic textures,Kalman filter,bag of systems,expectation maximization,sensitivity analysis,video annotation
Canopy clustering algorithm,Fuzzy clustering,Computer vision,CURE data clustering algorithm,Data stream clustering,Algorithm design,Pattern recognition,Correlation clustering,Computer science,Expectation–maximization algorithm,Artificial intelligence,Cluster analysis
Journal
Volume
Issue
ISSN
35
7
0162-8828
Citations 
PageRank 
References 
22
0.63
28
Authors
4
Name
Order
Citations
PageRank
Adeel Mumtaz1472.42
Emanuele Coviello279823.59
Gert R. G. Lanckriet34769296.98
Antoni B. Chan4165888.09