BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:A hierarchical model and a Bayesian strategy for unsupervised deco
 nvolution-segmentation of textured image
DTSTART:20190207T140000
DTEND:20190207T160000
DTSTAMP:20260407T042035Z
UID:0b2c6249969cf224601c856f1926b9f8931b0e173f9e59865f4cf394
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Jean-Francois GIOVANNELLI - Université de Bordeaux 1 (F
 rance)\nThe proposed talk deals with the problem of deconvolution-segmenta
 tion of textured images. It is specifically devoted to the case of oriente
 d textures and focused on unsupervised solutions. The images are composed 
 of patches of textures that belong to a set of K possible classes describe
 d by a Gaussian random field. The labels that define the patches are model
 led by a Potts field. The method relies on a hierarchical model and a Baye
 sian strategy to jointly estimate the labels\, the textured images as well
  as the hyperparameters including texture and the Potts parameters. The es
 timators are computed based on a convergent procedure\, from samples of th
 e posterior obtained through an MCMC algorithm (Gibbs sampler including Pe
 rturbation-Optimization and Fisher-oriented Metropolis-Hastings). A first 
 numerical evaluation is proposed. \n \n\nJean-François GIOVANNELLI was 
 born in Béziers\, France\, in 1966. He graduated from the Ecole National
 e Supérieure de l'Electronique et de ses Applications in 1990. He receiv
 ed the Ph. D. degree and the Habilitation à Diriger des Recherches in phy
 sics (signal-image processing) from the Université Paris-Sud\, in 1995 an
 d 2005 respectively. From 1997 to 2008\, he has been Assistant Professor w
 ith the Université Paris-Sud\, and a Researcher with the Laboratoire des 
 Signaux et Systèmes\, Groupe Problèmes Inverses. He is presently Profess
 or with the Université de Bordeaux and a Researcher with the Laboratoire 
 d'Intégration du Matériau au Système\, Groupe Signal-Image. He is inter
 ested in regularization and Bayesian methods for inverse problems in signa
 l and image processing\, mainly unsupervised and myopic (self-calibrated) 
 problems. His application fields essentially concern astronomical\, medica
 l\, proteomics and geophysical imaging. For more information: http://giov
 annelli.free.fr \n 
LOCATION:DIA 005 https://plan.epfl.ch/?room=DIA005
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR