Relaxation-Based Preprocessing Techniques for Markov Random Field Inference.
Chen WangRamin ZabihPublished in: CVPR (2016)
Keyphrases
- markov random field
- preprocessing
- random fields
- loopy belief propagation
- map inference
- mrf model
- belief propagation
- pairwise
- partition function
- higher order
- dual decomposition
- image segmentation
- parameter estimation
- graph cuts
- maximum a posteriori
- energy function
- energy minimization
- piecewise constant functions
- conditional random fields
- message passing
- probabilistic inference
- higher order cliques
- mrf models
- image restoration
- prior model
- neighboring pixels
- mrf optimization
- undirected graphical models
- structured prediction
- potential functions
- expectation maximization
- graphical models
- lagrangian relaxation
- bayesian networks
- feature extraction
- long range
- non stationary
- machine learning