High quality speech synthesis based on wavelet compilation of phoneme segments.
Tomohisa HirokawaKenzo ItohHirokazu SatoPublished in: ICSLP (1992)
Keyphrases
- speech synthesis
- high quality
- speech recognition
- text to speech
- vocal tract
- prosodic features
- wavelet transform
- multiscale
- multiresolution
- wavelet coefficients
- higher quality
- wavelet analysis
- low quality
- wavelet domain
- denoising
- ground truth
- image quality
- speech corpus
- wavelet transformation
- high fidelity
- multiresolution analysis
- high resolution
- pattern recognition
- image processing
- wavelet decomposition
- power spectrum
- original signal
- multi fractal
- inverse problems in image processing