Bayesian Singing Transcription Based on a Hierarchical Generative Model of Keys, Musical Notes, and F0 Trajectories.
Ryo NishikimiEita NakamuraMasataka GotoKatsutoshi ItoyamaKazuyoshi YoshiiPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2020)
Keyphrases
- generative model
- posterior probability
- pitman yor process
- music information retrieval
- prior distribution
- probabilistic model
- markov chain monte carlo
- polyphonic music
- bayesian framework
- mixture model
- dirichlet process mixture models
- prior knowledge
- latent dirichlet allocation
- fully bayesian
- discriminative learning
- hierarchical bayesian model
- semi supervised
- posterior distribution
- music retrieval
- bayesian networks
- topic models
- maximum likelihood
- discriminative models
- generative and discriminative models
- em algorithm
- moving objects
- variational inference
- bayesian inference
- clustering algorithm