Scaling MPE Inference for Constrained Continuous Markov Random Fields with Consensus Optimization.
Stephen H. BachMatthias BroechelerLise GetoorDianne P. O'LearyPublished in: NIPS (2012)
Keyphrases
- markov random field
- dual decomposition
- random fields
- loopy belief propagation
- map inference
- belief propagation
- belief networks
- graph cuts
- efficient inference
- partition function
- parameter estimation
- energy function
- probabilistic inference
- bayesian networks
- graphical models
- energy minimization
- higher order
- image segmentation
- piecewise constant functions
- maximum a posteriori
- parameter learning
- message passing
- potential functions
- conditional random fields
- pairwise
- mrf model
- markov networks
- image restoration
- image restoration and reconstruction
- low level vision
- exact inference
- factor graphs
- image labeling
- undirected graphical models
- non stationary
- potts model
- iterative conditional
- higher order cliques
- map estimation
- bayesian inference
- generalized belief propagation
- high level vision
- maximum entropy
- structured prediction
- textured images