Maximum marginal likelihood estimation of the granularity coefficient of a Potts-Markov random field within an MCMC algorithm.
Marcelo PereyraNick WhiteleyChristophe AndrieuJean-Yves TourneretPublished in: SSP (2014)
Keyphrases
- markov random field
- parameter estimation
- energy function
- mrf model
- loopy belief propagation
- graph cuts
- random fields
- energy minimization
- markov chain monte carlo
- em algorithm
- dynamic programming
- k means
- markov chain
- higher order
- probabilistic model
- monte carlo
- pairwise
- posterior probability
- image segmentation
- marginal likelihood
- hyperparameters
- conditional random fields
- expectation maximization
- state space
- bayesian networks