Variational Bayesian Image Modelling
We present a variational Bayesian framework
for performing inference, density estimation and
model selection in a special class of graphical
models—Hidden Markov Random Fields (HMRFs).
HMRFs are particularly well suited to image
modelling and in this paper, we apply them
to the problem of image segmentation. Unfortunately,
HMRFs are notoriously hard to train
and use because the exact inference problems
they create are intractable. Our main contribution
is to introduce an efficient variational approach
for performing approximate inference of
the Bayesian formulation of HMRFs, which we
can then apply to the density estimation and
model selection problems that arise when learning
image models from data. With this variational
approach, we can conveniently tackle the
problem of image segmentation. We present experimental
results which show that our technique
outperforms recent HMRF-based segmentation
methods on real world images.
Citation
L. Cheng,
F. Jiao,
D. Schuurmans,
S. Wang.
"Variational Bayesian Image Modelling".
International Conference on Machine Learning (ICML), Bonn, Germany, January 2005.
Keywords: |
variational, machine learning |
Category: |
In Conference |
BibTeX
@incollection{Cheng+al:ICML05,
author = {Li Cheng and Feng Jiao and Dale Schuurmans and Shaojun Wang},
title = {Variational Bayesian Image Modelling},
booktitle = {International Conference on Machine Learning (ICML)},
year = 2005,
}
Last Updated: April 25, 2007
Submitted by William Thorne