Markov Random Field Energy Minimization via Iterated Cross Entropy with Partition Strategy.
Jue WuAlbert C. S. ChungPublished in: ICASSP (1) (2007)
Keyphrases
- energy minimization
- markov random field
- cross entropy
- energy function
- graph cuts
- belief propagation
- min cut
- random fields
- mrf model
- parameter estimation
- low level vision
- higher order
- maximum likelihood
- pairwise
- image segmentation
- log likelihood
- max flow
- image restoration
- conditional random fields
- maximum a posteriori
- active contour model
- potential functions
- dual decomposition
- global minimum
- evaluation metrics
- iterated conditional modes
- energy minimization problems
- map estimation
- mrf models
- discontinuity preserving
- map inference
- potts model
- loopy belief propagation
- information retrieval
- message passing
- multiscale
- similarity measure