Abstract

Probabilistic models have been adopted for many computer vision applications, however inference in highdimensional spaces remains problematic. As the statespace of a model grows, the dependencies between the dimensions lead to an exponential growth in computation when performing inference. Many common computer vision problems naturally map onto the graphical model framework; the representation is a graph where each node contains a portion of the state-space and there is an edge between two nodes only if they are not independent conditional on the other nodes in the graph. When this graph is sparsely connected, belief propagation algorithms can turn an exponential inference computation into one which is linear in the size of the graph. However belief propagation is only applicable when the variables in the nodes are discrete-valued or jointly represented by a single multivariate Gaussian distribution, and this rules out many computer vision applications.

Citations

3012	Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images - Geman, Geman - 1984
2438	Snakes: active contour models - Kass, Witkin, et al. - 1988
911	Condensation - conditional density propagation for visual tracking - Isard, Blake - 1998
382	Learning lowlevel vision - Freeman, Pasztor, et al.
228	Tractable inference for complex stochastic processes - Boyen, Koller - 1998
214	Understanding belief propagation and its generalizations - Yedidia, Freeman, et al. - 2001
185	A family of algorithms for approximate Bayesian inference - Minka - 2001
142	Stochastic completion fields: A neural model of illusory contour shape and salience - Williams, Jacobs - 1997
139	Nonparametric belief propagation - Sudderth, Ihler, et al. - 2003
131	Implicit probabilistic models of human motion for synthesis and tracking - Sidenbladh, Black, et al. - 2002
114	Efficient matching of pictorial structures - Felzenszwalb, Huttenlocher - 2000
106	Hierarchical Bayesian inference in the visual cortex - Lee, Mumford - 2003
104	Estimating human body configurations using shape context matching - Mori, Malik - 2002
74	Finding deformable shapes using loopy belief propagation - Coughlan, Ferreira - 2002
62	A general algorithm for approximate inference and its application to hybrid bayes nets - Koller, Lerner, et al. - 1999
46	Human tracking with mixtures of trees - Ioffe, Forsyth - 2001
45	A smoothing filter for condensation - Isard, Blake - 1998
39	Jetstream: Probabilistic contour extraction with particles - PREZ, BLAKE, et al. - 2001
38	Transformed Hidden Markov Models: Estimating Mixture Models and Inferring Spatial Transformations - Jojic, Petrovic, et al. - 2000
37	Hybrid Bayesian Networks for Reasoning About Complex Systems - Lerner - 2002
29	VIBES: A variational inference engine for Bayesian networks - Bishop, Spiegelhalter, et al. - 2003
10	Mixed Markov models - Fridman - 2003
8	The Curve Indicator Random Field: Curve Organization Via Edge Correlation,” Perceptual Organization for Artificial Vision Systems - August, Zucker - 2000