Exact Optimization for Markov Random Fields with Convex Priors.
Hiroshi IshikawaPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2003)
Keyphrases
- markov random field
- maximum a posteriori
- labeling problems
- graph cuts
- prior model
- map estimation
- parameter estimation
- random fields
- higher order
- energy function
- belief propagation
- energy minimization
- mrf model
- pairwise
- image restoration
- image restoration and reconstruction
- image segmentation
- prior models
- potential functions
- low level vision
- conditional random fields
- higher order cliques
- convex relaxation
- image prior
- convex optimization
- map inference
- efficient inference
- iterative conditional
- loopy belief propagation
- piecewise constant functions
- potts model
- bayesian framework
- convex programming
- prior knowledge
- markov networks
- stereo matching
- textured images
- maximum likelihood
- smoothness prior
- alternating direction method of multipliers
- computer vision
- message passing