by Mark Berthod, Zoltan Kato, Josiane Zerubia
Abstract:
Deterministic pseudo-annealing (DPA) is a new deterministic optimization method for finding the maximum a posteriori (MAP) labeling in a Markov random field, in which the probability of a tentative labeling is extended to a merit function on continuous labelings. This function is made convex by changing its definition domain. This unambiguous maximization problem is solved, and the solution is followed down to the original domain, yielding a good, if suboptimal, solution to the original labeling assignment problem. The performance of DPA is analyzed on randomly weighted graphs
Reference:
Mark Berthod, Zoltan Kato, Josiane Zerubia, DPA: A Deterministic Approach to the MAP, In IEEE Transactions on Image Processing, volume 4, no. 9, pp. 1312-1314, 1995.
Bibtex Entry:
@string{tip="IEEE Transactions on Image Processing"}
@Article{Berthod-etal95c,
author = {Berthod, Mark and Kato, Zoltan and Zerubia, Josiane},
title = {{DPA}: A Deterministic Approach to the {MAP}},
journal = tip,
year = 1995,
volume = 4,
number = 9,
pages = {1312--1314},
month = sep,
abstract = {Deterministic pseudo-annealing (DPA) is a new
deterministic optimization method for finding the
maximum a posteriori (MAP) labeling in a Markov
random field, in which the probability of a
tentative labeling is extended to a merit function
on continuous labelings. This function is made
convex by changing its definition domain. This
unambiguous maximization problem is solved, and the
solution is followed down to the original domain,
yielding a good, if suboptimal, solution to the
original labeling assignment problem. The
performance of DPA is analyzed on randomly weighted
graphs}
}