A combination approach of Gaussian mixture models and support vector machines for speaker identification.
Rafik DjemiliHocine BouroubaMohamed Cherif Amara KorbaPublished in: Int. Arab J. Inf. Technol. (2009)
Keyphrases
- gaussian mixture model
- speaker identification
- support vector
- speaker recognition
- mixture model
- language identification
- probabilistic neural network
- feature vectors
- em algorithm
- mel frequency cepstral coefficients
- expectation maximization
- maximum likelihood
- noisy environments
- feature space
- neural network
- classification accuracy
- broadcast news
- probabilistic model
- speech recognition
- prior knowledge
- object recognition
- feature extraction
- image processing
- machine learning