A Maximum Likelihood Estimation of Vocal-Tract-Related Filter Characteristics for Single Channel Speech Separation.
Mohammad H. RadfarRichard M. DansereauAbolghasem SayadiyanPublished in: EURASIP J. Audio Speech Music. Process. (2007)
Keyphrases
- speech enhancement
- maximum likelihood estimation
- vocal tract
- single channel
- noise reduction
- speech signal
- speech synthesis
- wiener filter
- multi channel
- noisy environments
- sound source
- maximum likelihood
- speech recognition
- image acquisition
- em algorithm
- signal to noise ratio
- expectation maximization
- probability distribution
- linear prediction
- speech sounds
- image restoration
- independent component analysis
- multi modal
- brain activity
- text to speech
- frequency domain
- signal processing
- hidden markov models
- multiscale
- prior information
- higher order
- pattern recognition
- machine learning