Reducing the Computational Cost of Deep Generative Models with Binary Neural Networks.
Thomas BirdFriso H. KingmaDavid BarberPublished in: ICLR (2021)
Keyphrases
- generative model
- neural network
- probabilistic model
- mixture model
- discriminative learning
- discriminative models
- semi supervised
- em algorithm
- deep belief networks
- generative and discriminative models
- conditional random fields
- prior knowledge
- hidden variables
- hierarchical hidden markov models
- maximum entropy principle
- pairwise
- representational power
- topic models
- object categories
- unsupervised learning
- artificial neural networks
- object recognition