A Mixture Model Based Markov Random Field for Discovering Patterns in Sequences.
Konstantinos BlekasPublished in: SETN (2006)
Keyphrases
- markov random field
- discovering patterns
- graph cuts
- mrf model
- data mining
- pattern discovery
- belief propagation
- higher order
- pairwise
- maximum a posteriori
- parameter estimation
- energy function
- energy minimization
- random fields
- image restoration
- hidden markov models
- neighboring pixels
- image segmentation
- conditional random fields
- message passing
- sequential patterns
- prior model
- mrf optimization
- mrf models
- loopy belief propagation
- textured images
- spatial continuity
- machine learning
- markov random field model
- real world
- map estimation
- potential functions
- random walk
- mixture model
- multi dimensional
- multiscale
- simulated annealing
- active contours
- bayesian framework