Global variance modeling on the log power spectrum of LSPs for HMM-based speech synthesis.
Zhen-Hua LingYu HuLi-Rong DaiPublished in: INTERSPEECH (2010)
Keyphrases
- power spectrum
- speech synthesis
- frequency domain
- natural images
- speech recognition
- frequency analysis
- hidden markov models
- text to speech
- power spectral
- spectral estimation
- vocal tract
- higher order
- object recognition
- autocorrelation function
- signal noise ratio
- prosodic features
- multiresolution
- similarity measure
- neural network