Improving phoneme and accent estimation by leveraging a dictionary for a stochastic TTS front-end.
Tohru NaganoRyuki TachibanaNobuyasu ItohMasafumi NishimuraPublished in: ICASSP (2008)
Keyphrases
- speech recognition
- speech synthesis
- automatic speech recognition
- text to speech
- prosodic features
- spoken language
- estimation accuracy
- estimation algorithm
- back end
- monte carlo
- sparse representation
- language model
- stochastic nature
- real time
- context dependent
- stochastic model
- parameter estimation
- learning automata
- stochastic optimization
- estimation process
- image classification
- least squares