Synthesis of stressed speech from isolated neutral speech using HMM-based models.
Sahar E. Bou-GhazaleJohn H. L. HansenPublished in: ICSLP (1996)
Keyphrases
- speech recognition
- hidden markov models
- facial animation
- speech signal
- probabilistic model
- speech synthesis
- neural network
- speech processing
- speaker identification
- broadcast news
- automatic speech recognition
- recognition engine
- statistical models
- experimental data
- complex systems
- model selection
- acoustic models
- audio stream
- automatic speech recognition systems
- text to speech
- noisy environments
- audio visual
- autoregressive
- statistical model
- prior knowledge