Efficient hardware acceleration for approximate inference of bitwise deep neural networks.
Sebastian VogelAndre GuntoroGerd AscheidPublished in: DASIP (2017)
Keyphrases
- approximate inference
- neural network
- graphical models
- probabilistic inference
- message passing
- belief propagation
- variational methods
- parameter estimation
- factor graphs
- exact inference
- high quality
- information retrieval
- gaussian process
- latent variables
- conditional random fields
- random variables
- markov random field
- dynamic bayesian networks
- artificial neural networks
- training data
- loopy belief propagation
- image processing
- expectation propagation