Learned Decimation for Neural Belief Propagation Decoders.
Andreas BuchbergerChristian HägerHenry D. PfisterLaurent SchmalenAlexandre Graell i AmatPublished in: CoRR (2020)
Keyphrases
- belief propagation
- graphical models
- message passing
- stereo matching
- decoding algorithm
- graph cuts
- markov random field
- approximate inference
- pairwise
- factor graphs
- neural network
- loopy belief propagation
- markov networks
- free energy
- fixed point
- energy minimization
- statistical physics
- stereo vision
- energy function
- belief networks
- probabilistic inference
- exact inference
- probabilistic model
- parameter estimation
- maximum likelihood
- probability distribution
- dynamic programming
- feature extraction
- image sequences
- three dimensional