Combining Gaussian Mixture Model with Global Variance Term to Improve the Quality of an HMM-Based Polyglot Speech Synthesizer.
Javier LatorreKoji IwanoSadaoki FuruiPublished in: ICASSP (4) (2007)
Keyphrases
- gaussian mixture model
- speaker recognition
- speaker identification
- mixture model
- text to speech
- hidden markov models
- maximum likelihood
- expectation maximization
- em algorithm
- feature space
- gaussian distribution
- background subtraction
- machine learning
- feature vectors
- automatic speech recognition
- speech recognition
- gaussian mixture modeling
- probability density function
- standard deviation
- gaussian mixture
- gaussian model
- image quality
- mixture distribution