Low Noise Non-Linear Equalization Using Neural Networks and Belief Propagation.
Etsushi YamazakiNariman FarsadAndrea GoldsmithPublished in: CoRR (2019)
Keyphrases
- belief propagation
- neural network
- graphical models
- message passing
- markov random field
- graph cuts
- stereo matching
- approximate inference
- fixed point
- markov networks
- loopy belief propagation
- pairwise
- factor graphs
- energy function
- competitive learning
- energy minimization
- noise level
- random variables
- self organizing maps
- exact inference
- free energy
- sum product
- maximum a posteriori
- probabilistic model
- high quality
- image segmentation
- stereo vision
- image matching
- post processing