Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

DPA: a deterministic approach to the MAP problem.

M Berthod1, Z Kato, J Zerubia

  • 1Inst. Nat. de Recherche en Inf. et Autom., Sophia Antipolis.

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|January 1, 1995
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A locally adaptive peano scanning algorithm.

IEEE transactions on pattern analysis and machine intelligence·2011
Same author

Improving consistency and reducing ambiguity in stochastic labeling: an optimization approach.

IEEE transactions on pattern analysis and machine intelligence·2011
Same author

Building Development Monitoring in Multitemporal Remotely Sensed Image Pairs with Stochastic Birth-Death Dynamics.

IEEE transactions on pattern analysis and machine intelligence·2011
Same author

Antigen-specific immune response to endotoxin-free recombinant P34.

Allergy·2011
Same author

Clinical and radiological features of rotavirus cerebellitis.

AJNR. American journal of neuroradiology·2010
Same author

Scoring clinical functions in subacute sclerosing panencephalitis.

AJNR. American journal of neuroradiology·2009
Same journal

Change-Prior-Guided Unsupervised Change Detection of Heterogeneous Remote Sensing Images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

AgonicDreamer: Enhancing Multi-View Consistency in Text-to-3D Generation via Rectified Score Distillation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

BiCM-Prompt: Bidirectional Cross-Modal Prompt Tuning for Class-Incremental Learning on Multisource Remote Sensing Images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

GoP-based Quality Enhancement on Video Compression.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Align then Tensorize: Multi-Level Consistent Anchor Graph Learning for Scalable Multi-View Clustering.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Beyond Fidelity: Diverse Image Synthesis via Retrieval-Augmented Diffusion.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
See all related articles

Deterministic pseudo-annealing (DPA) offers a novel optimization method for Markov random fields. This approach finds maximum a posteriori (MAP) labelings by solving a convex continuous problem, providing effective solutions for labeling assignment.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Markov random fields are widely used for modeling complex systems.
  • Finding the maximum a posteriori (MAP) labeling is a critical but computationally challenging problem.

Purpose of the Study:

  • Introduce Deterministic Pseudo-Annealing (DPA) as a new deterministic optimization method.
  • Address the challenge of efficiently finding MAP labelings in Markov random fields.

Main Methods:

  • Extend tentative labeling probabilities to a merit function on continuous labelings.
  • Transform the merit function into a convex problem by altering its domain.
  • Solve the convex maximization problem and trace the solution back to the original domain.

Related Experiment Videos

Main Results:

  • DPA provides a deterministic approach to MAP labeling.
  • The method yields good, albeit potentially suboptimal, solutions for labeling assignment problems.
  • Performance analysis was conducted on randomly weighted graphs.

Conclusions:

  • Deterministic pseudo-annealing is a viable optimization technique for Markov random fields.
  • The method offers a practical approach to solving the MAP labeling problem.
  • Further analysis on various graph structures can validate its broader applicability.