Modulation spectrum compensation for HMM-based speech synthesis using line spectral pairs.
Zhen-Hua LingXiao-Hui SunLi-Rong DaiYu HuPublished in: ICASSP (2016)
Keyphrases
- speech synthesis
- speech recognition
- hidden markov models
- spectral resolution
- text to speech
- tandem mass
- prosodic features
- vocal tract
- pairwise
- line segments
- short wave
- frequency spectrum
- hyperspectral
- hyperspectral imagery
- multispectral
- infrared
- field of view
- automatic speech recognition
- frequency domain
- computer simulation
- language model
- image segmentation