Unsupervised vocal dereverberation with diffusion-based generative models.
Koichi SaitoNaoki MurataToshimitsu UesakaChieh-Hsin LaiYuhta TakidaTakao FukuiYuki MitsufujiPublished in: CoRR (2022)
Keyphrases
- generative model
- semi supervised
- probabilistic model
- discriminative learning
- mixture model
- discriminative models
- unsupervised learning
- hierarchical hidden markov models
- deep belief networks
- latent dirichlet allocation
- prior knowledge
- semi supervised learning
- conditional random fields
- hidden variables
- supervised learning
- generative and discriminative models
- em algorithm
- topic models
- object categories
- labeled data
- maximum entropy principle
- weakly supervised
- text classification
- restricted boltzmann machine
- active learning
- support vector