Discriminative speaker recognition using large margin GMM.
Reda JouraniKhalid DaoudiRégine André-ObrechtDriss AboutajdinePublished in: Neural Comput. Appl. (2013)
Keyphrases
- speaker recognition
- speaker identification
- gaussian mixture model
- discriminative training
- discriminative learning
- speech recognition
- feature transformation
- automatic speech recognition
- feature extraction
- speaker verification
- maximum likelihood
- hidden markov models
- gaussian mixture modeling
- speech signal
- mel frequency cepstral coefficients
- feature vectors
- vector quantization
- linear classifiers
- broadcast news
- speaker diarization
- discriminative power
- novelty detection
- gaussian mixture
- mixture model
- labeled data
- semi supervised
- discriminative features
- probabilistic neural network
- hyperplane
- structured learning
- em algorithm
- weakly supervised
- maximum margin
- audio visual
- probabilistic interpretation
- unsupervised learning
- support vector