A family of Bayesian STSA estimators for the enhancement of speech with correlated frequency components.
Eric PlourdeBenoît ChampagnePublished in: ICASSP (2010)
Keyphrases
- speech recognition
- recognition engine
- bayesian networks
- image enhancement
- special case
- speech synthesis
- speech signal
- data driven
- bayesian inference
- bayesian learning
- image processing
- highly correlated
- analytical expressions
- text to speech
- gaussian processes
- posterior probability
- neural network
- bayesian methods
- dialogue system
- bayesian decision
- pattern recognition
- language acquisition
- linear prediction
- spoken language
- audio visual
- maximum likelihood
- asymptotic properties
- variance estimator