Sparse Gaussian Process Audio Source Separation Using Spectrum Priors in the Time-domain.
Pablo A. AlvaradoMauricio A. ÁlvarezDan StowellPublished in: ICASSP (2019)
Keyphrases
- gaussian process
- source separation
- bayesian framework
- gaussian processes
- hyperparameters
- regression model
- audio features
- multimedia
- latent variables
- prior information
- model selection
- frequency domain
- prior knowledge
- sparse representation
- semi supervised
- high dimensional
- generative model
- audio visual
- machine learning
- signal processing
- hidden markov models
- maximum a posteriori
- text mining
- multiresolution
- blind source separation
- dimensionality reduction