Low complexity sparse Bayesian learning using combined belief propagation and mean field with a stretched factor graph.
Chuanzong ZhangZhengdao YuanZhongyong WangQinghua GuoPublished in: Signal Process. (2017)
Keyphrases
- belief propagation
- factor graphs
- markov random field
- markov networks
- loopy belief propagation
- graphical models
- message passing
- free energy
- approximate inference
- exact inference
- belief networks
- graph cuts
- stereo matching
- partition function
- probabilistic graphical models
- computational complexity
- energy function
- higher order
- probabilistic model
- pairwise
- motion estimation
- conditional random fields
- parameter estimation
- maximum a posteriori
- image segmentation
- latent variables
- bayesian networks
- bayesian inference
- maximum margin
- variational methods
- structured prediction
- random fields
- probabilistic inference
- approximation algorithms
- closed form
- hidden markov models