Blind source separation from noisy data using bayesian estimation and gibbs priors.
Luigi BediniEmanuele SalernoAnna TonazziniPublished in: SIP (2001)
Keyphrases
- bayesian estimation
- noisy data
- blind source separation
- maximum a posteriori
- posterior distribution
- independent component analysis
- bayesian framework
- non stationary
- markov random field
- negative matrix factorization
- map estimation
- missing data
- missing values
- hyperparameters
- high dimensional
- training data
- input data
- prior knowledge
- prior information
- sparse representation
- speech signal
- learning algorithm
- least squares
- posterior probability