Detail publikačního výsledku

Fast Bayesian Estimation Using Location-type Variational Representations

DORAZIL, J.; HLAWATSCH, F.; FLEURY, B.; BURGET, R.

Originální název

Fast Bayesian Estimation Using Location-type Variational Representations

Anglický název

Fast Bayesian Estimation Using Location-type Variational Representations

Druh

Stať ve sborníku v databázi WoS či Scopus

Originální abstrakt

A class of iterative Bayesian estimation methods known as Type I methods use a variational representation of the posterior distribution. However, Type I methods can exhibit slow convergence for a family of variational representations referred to as convex/location representation. We analyze the convergence behavior of Type I methods by interpreting them as an iterative maximization of a dual objective function involving a linearization. We then propose a modified method that avoids the linearization and allows a coordinatewise maximization. We demonstrate the advantages of the proposed method in the context of image restoration under Poisson noise.

Anglický abstrakt

A class of iterative Bayesian estimation methods known as Type I methods use a variational representation of the posterior distribution. However, Type I methods can exhibit slow convergence for a family of variational representations referred to as convex/location representation. We analyze the convergence behavior of Type I methods by interpreting them as an iterative maximization of a dual objective function involving a linearization. We then propose a modified method that avoids the linearization and allows a coordinatewise maximization. We demonstrate the advantages of the proposed method in the context of image restoration under Poisson noise.

Klíčová slova

Bayesian estimation; Type I estimation; half-quadratic minimization; variational representation; convex representation, location parameterization; image restoration

Klíčová slova v angličtině

Bayesian estimation; Type I estimation; half-quadratic minimization; variational representation; convex representation, location parameterization; image restoration

Autoři

DORAZIL, J.; HLAWATSCH, F.; FLEURY, B.; BURGET, R.

Rok RIV

2026

Vydáno

09.09.2025

ISBN

978-9-46-459362-4

Kniha

Procecedings of the 33rd European Signal Processing Conference EUSIPCO 2025

Strany od

2427

Strany do

2431

Strany počet

5

URL

https://eurasip.org/Proceedings/Eusipco/Eusipco2025/pdfs/0002427.pdf

BibTex

@inproceedings{BUT199983,
  author="Jan {Dorazil} and Franz {Hlawatsch} and Bernard H. {Fleury} and Radim {Burget}",
  title="Fast Bayesian Estimation Using Location-type Variational Representations",
  booktitle="Procecedings of the 33rd European Signal Processing Conference
EUSIPCO 2025",
  year="2025",
  pages="2427--2431",
  isbn="978-9-46-459362-4",
  url="https://eurasip.org/Proceedings/Eusipco/Eusipco2025/pdfs/0002427.pdf"
}