by Josiane Zerubia, Zoltan Kato, Mark Berthod
Abstract:
As it is well known, optimization of the energy function of Markov Random Fields is very expensive. Hierarchical models have usually much more communication per pixel than monogrid ones. This is why classical annealing schemes are too slow, even on a parallel machine, to minimize the energy associated with such a model. However, taking benefit of the pyramidal structure of the model, we can define a new annealing scheme: the Multi-Temperature Annealing (MTA), which consists of associating higher temperatures to coarser levels, in order to be less sensitive to local minima at coarser grids. The convergence to the global optimum is proved by a generalisation of the annealing theorem of Geman and Geman. We have applied the algorithm to image classification and tested it on synthetic and real images.
Reference:
Josiane Zerubia, Zoltan Kato, Mark Berthod, Multi-Temperature Annealing: A New Approach for the Energy-Minimization of Hierarchical Markov Random Field Models, In Proceedings of International Conference on Pattern Recognition, volume 1, Jerusalem, Israel, pp. 520-522, 1994, IEEE.
Bibtex Entry:
@string{icpr="Proceedings of International Conference on Pattern Recognition"}
@InProceedings{Zerubia-etal94,
author = {Zerubia, Josiane and Kato, Zoltan and Berthod, Mark},
title = {Multi-Temperature Annealing: A New Approach for the
Energy-Minimization of Hierarchical {M}arkov Random
Field Models},
booktitle = icpr,
pages = {520-522},
year = 1994,
volume = 1,
address = {Jerusalem, Israel},
month = oct,
organization = {IAPR},
publisher = {IEEE},
ps = {../papers/icpr94.ps},
abstract = {As it is well known, optimization of the energy
function of Markov Random Fields is very
expensive. Hierarchical models have usually much
more communication per pixel than monogrid
ones. This is why classical annealing schemes are
too slow, even on a parallel machine, to minimize
the energy associated with such a model. However,
taking benefit of the pyramidal structure of the
model, we can define a new annealing scheme: the
Multi-Temperature Annealing (MTA), which consists of
associating higher temperatures to coarser levels,
in order to be less sensitive to local minima at
coarser grids. The convergence to the global optimum
is proved by a generalisation of the annealing
theorem of Geman and Geman. We have applied the
algorithm to image classification and tested it on
synthetic and real images.}
}