Adaptation of pitch and spectrum for HMM-based speech synthesis using MLLR.
Masatsune TamuraTakashi MasukoKeiichi TokudaTakao KobayashiPublished in: ICASSP (2001)
Keyphrases
- speech synthesis
- speaker adaptation
- speech recognition
- hidden markov models
- vocal tract
- speaker independent
- automatic speech recognition
- speech recognizer
- speech signal
- prosodic features
- language model
- image processing
- speech recognition systems
- text to speech
- noisy environments
- speech corpus
- acoustic features
- variable duration
- pattern recognition
- multi fractal
- data sets
- formant frequencies
- case based reasoning
- website
- knowledge base