Auxiliary variables in conditional Gaussian mixtures for automatic speech recognition.
Todd A. StephensonMathew Magimai-DossHervé BourlardPublished in: INTERSPEECH (2002)
Keyphrases
- automatic speech recognition
- gaussian mixture
- auxiliary variables
- speech recognition
- speech signal
- em algorithm
- deformable models
- gaussian mixture model
- hidden markov models
- higher order
- expectation maximization
- covariance matrix
- graph cuts
- half quadratic
- probability density function
- mixture model
- closed form
- conversational speech
- density estimation
- pattern recognition
- image analysis
- image segmentation
- maximum likelihood
- natural images