Linear and non linear kernel GMM supervector machines for speaker verification.
Réda DehakNajim DehakPatrick KennyPierre DumouchelPublished in: INTERSPEECH (2007)
Keyphrases
- gaussian mixture model
- speaker verification
- speaker recognition
- noisy environments
- feature space
- speaker identification
- mixture model
- audio visual
- em algorithm
- prosodic features
- language identification
- maximum likelihood
- support vector
- expectation maximization
- feature vectors
- acoustic features
- emotion recognition
- multilayer perceptron