A Convex Discrete-Continuous Approach for Markov Random Fields.
Christopher ZachPushmeet KohliPublished in: ECCV (6) (2012)
Keyphrases
- markov random field
- discrete space
- graph cuts
- continuous domains
- belief propagation
- mrf model
- image restoration
- maximum a posteriori
- continuous variables
- energy function
- parameter estimation
- random fields
- higher order
- energy minimization
- convex sets
- labeling problems
- image segmentation
- message passing
- convex optimization
- conditional random fields
- pairwise
- low level vision
- efficient inference
- piecewise constant functions
- discriminative random fields
- iterative conditional
- map inference
- loopy belief propagation
- potential functions
- partition function
- min cut
- image processing
- potts model
- object recognition
- convex programming
- textured images
- dynamic programming
- convex hull