DVAE#: Discrete Variational Autoencoders with Relaxed Boltzmann Priors.

Arash Vahdat Evgeny Andriyash William G. Macready

Published in: NeurIPS (2018)

Keyphrases

denoising
discrete geometry
data sets
image processing
image segmentation
prior knowledge
discrete version
optical flow
unsupervised learning
prior information
bayesian framework
continuous variables
discrete space
optical flow computation
learned from training data
framework for image segmentation