Embedded kernel eigenvoice speaker adaptation and its implication to reference speaker weighting.
Brian Kan-Wing MakRoger Wend-Huu HsiaoSimon Ka-Lung HoJames T. KwokPublished in: IEEE Trans. Speech Audio Process. (2006)
Keyphrases
- speaker adaptation
- speech recognition
- maximum likelihood
- speaker dependent
- automatic speech recognition
- speech recognizer
- speaker independent
- hidden markov models
- feature space
- language model
- speaker identification
- principal component analysis
- multi modal
- non stationary
- spoken language
- broadcast news
- pattern recognition