Comparison of subspace methods for Gaussian mixture models in speech recognition.
Matti VarjokallioMikko KurimoPublished in: INTERSPEECH (2007)
Keyphrases
- speech recognition
- gaussian mixture model
- speaker identification
- speaker recognition
- subspace methods
- mixture model
- language model
- hidden markov models
- mel frequency cepstral coefficients
- automatic speech recognition
- speech signal
- em algorithm
- expectation maximization
- speech synthesis
- maximum likelihood
- feature vectors
- noisy environments
- probabilistic neural network
- feature space
- speech recognition systems
- pattern recognition
- ensemble methods
- vector quantization
- probabilistic model
- face recognition
- linear discriminant analysis
- speaker verification
- unsupervised learning
- generative model
- cepstral coefficients