Stochastic Modeling and Quantization of Harmonic Phases in Speech using Wrapped Gaussian Mixture Models.
Yannis AgiomyrgiannakisYannis StylianouPublished in: ICASSP (4) (2007)
Keyphrases
- gaussian mixture model
- speaker recognition
- speaker identification
- mixture model
- speech music discrimination
- feature vectors
- speech recognition
- em algorithm
- speech signal
- mel frequency cepstral coefficients
- expectation maximization
- gaussian mixture
- maximum likelihood
- covariance matrices
- probability density function
- density estimation
- gaussian distribution
- speaker verification
- multi modal
- probabilistic neural network
- object recognition
- linear predictive
- mixture modeling
- variational bayes
- probabilistic model
- noisy environments