DPA: A Deterministic Approach to the MAP Problem

Marc Berthod, Z. Kato, Josiane Zerubia

Research output: Contribution to journalArticle

8 Citations (Scopus)

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 convvex 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.

Original languageEnglish
Pages (from-to)1312-1314
Number of pages3
JournalIEEE Transactions on Image Processing
Volume4
Issue number9
DOIs
Publication statusPublished - 1995

Fingerprint

Maximum a Posteriori
Annealing
Labeling
Merit Function
Weighted Graph
Assignment Problem
Random Field
Optimization Methods

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Software
  • Computational Theory and Mathematics
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Theoretical Computer Science

Cite this

DPA : A Deterministic Approach to the MAP Problem. / Berthod, Marc; Kato, Z.; Zerubia, Josiane.

In: IEEE Transactions on Image Processing, Vol. 4, No. 9, 1995, p. 1312-1314.

Research output: Contribution to journalArticle

Berthod, Marc ; Kato, Z. ; Zerubia, Josiane. / DPA : A Deterministic Approach to the MAP Problem. In: IEEE Transactions on Image Processing. 1995 ; Vol. 4, No. 9. pp. 1312-1314.
@article{38dc8cb599e548faabb9faafd5a1166f,
title = "DPA: A Deterministic Approach to the MAP Problem",
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 convvex 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.",
author = "Marc Berthod and Z. Kato and Josiane Zerubia",
year = "1995",
doi = "10.1109/83.413175",
language = "English",
volume = "4",
pages = "1312--1314",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "9",

}

TY - JOUR

T1 - DPA

T2 - A Deterministic Approach to the MAP Problem

AU - Berthod, Marc

AU - Kato, Z.

AU - Zerubia, Josiane

PY - 1995

Y1 - 1995

N2 - 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 convvex 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.

AB - 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 convvex 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.

UR - http://www.scopus.com/inward/record.url?scp=0029375669&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029375669&partnerID=8YFLogxK

U2 - 10.1109/83.413175

DO - 10.1109/83.413175

M3 - Article

AN - SCOPUS:0029375669

VL - 4

SP - 1312

EP - 1314

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 9

ER -