Speaker adaptation for context-dependent HMM using spatial relation of both phoneme context hierarchy and speakers.
Yasuhiro KomoriTetsuo KosakaMasayuki YamadaHiroki YamamotoPublished in: EUROSPEECH (1997)
Keyphrases
- context dependent
- speaker adaptation
- speech recognition
- spatial relations
- maximum likelihood
- automatic speech recognition
- speaker dependent
- hidden markov models
- speaker independent
- low level
- speech recognizer
- natural language
- pattern recognition
- high level
- vocal tract
- speech synthesis
- closely related
- spatial information
- computer vision
- speech signal
- data mining
- image retrieval
- bayesian networks
- similarity measure