Improving speaker recognition using phonetically structured Gaussian mixture models.
Robert FaltlhauserGünther RuskePublished in: INTERSPEECH (2001)
Keyphrases
- gaussian mixture model
- speaker recognition
- speaker identification
- mixture model
- em algorithm
- feature vectors
- speaker verification
- feature space
- maximum likelihood
- mel frequency cepstral coefficients
- probability density function
- probabilistic neural network
- expectation maximization
- probability density
- gaussian mixture
- covariance matrices
- gaussian distribution
- density estimation
- pattern classification
- vector quantization
- bayesian inference
- language identification
- speech recognition
- finite mixtures