Markov Random Fields Can Bridge Levels of Abstraction.
Paul R. CooperPeter N. ProkopowiczPublished in: NIPS (1991)
Keyphrases
- levels of abstraction
- markov random field
- graph cuts
- mrf model
- belief propagation
- energy function
- higher order
- image restoration
- maximum a posteriori
- image segmentation
- parameter estimation
- random fields
- low level vision
- pairwise
- energy minimization
- conditional random fields
- potential functions
- abstraction levels
- textured images
- message passing
- discriminative random fields
- map inference
- iterative conditional
- image processing
- potts model
- natural images
- map estimation
- efficient inference
- denoising
- object recognition
- markov networks
- loopy belief propagation
- least squares
- computer vision
- piecewise constant functions