A constrained sequential EM algorithm for speech enhancement.
Sunho ParkSeungjin ChoiPublished in: Neural Networks (2008)
Keyphrases
- em algorithm
- speech enhancement
- expectation maximization
- noise reduction
- noisy environments
- signal to noise ratio
- maximum likelihood
- mixture model
- single channel
- parameter estimation
- generative model
- maximum likelihood estimation
- gaussian mixture model
- speech signal
- expectation maximisation
- linear prediction
- incomplete data
- hyperparameters
- multi channel
- probability density function
- multi modal
- vocal tract
- smoothing algorithm
- probabilistic model
- image coding
- speech recognition
- machine learning