Non parametric Bayesian priors for hidden Markov random fields: application to image segmentation - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2018

Non parametric Bayesian priors for hidden Markov random fields: application to image segmentation

Résumé

One of the central issues in statistics and machine  learning is how to select an adequate model that can automatically adapt its complexity to the observed data. Bayesian nonparametric methods are thought of as one of the most promising candidates that  are capable of handling such tasks.  In the present work, we consider the issue of determining from the data the number of groups in a clustering task of non independent observations. The required guess on the number of  clusters is avoided by considering models with an infinite number of clusters as suggested in Dirichlet Process Mixture models (DPMM). However, for tasks such as unsupervised image segmentation with spatial relationships or dependencies between the observations, DPMM are not satisfying. We propose  to combine a Markov random field model with  different Bayesian nonparametric priors and illustrate such a combination on a Potts model combined with a Dirichlet process. As regards inference,  the variational expectation-maximization algorithm is adopted due  to its lower computational cost with respect to its MCMC counterpart. Finally, the proposed framework  is applied to  image segmentation  and some preliminary results are presented.  
Fichier non déposé

Dates et versions

hal-01941687 , version 1 (01-12-2018)

Identifiants

  • HAL Id : hal-01941687 , version 1

Citer

Florence Forbes, Hongliang Lu, Julyan Arbel. Non parametric Bayesian priors for hidden Markov random fields: application to image segmentation. BNPSI 2018 : Workshop on Bayesian non parametrics for signal and image processing, Jul 2018, Bordeaux, France. ⟨hal-01941687⟩
91 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More