High Quality Sinusoidal Modeling of Wideband Speech for the Purposes of Speech Synthesis and Modification.
Dan ChazanRon HooryAriel SagiSlava ShechtmanAlexander SorinZhiwei ShuangRaimo BakisPublished in: ICASSP (1) (2006)
Keyphrases
- speech synthesis
- speech recognition
- high quality
- text to speech
- vocal tract
- speech signal
- prosodic features
- speech corpus
- image quality
- speaker identification
- automatic speech recognition
- spontaneous speech
- speaker diarization
- noisy environments
- modeling method
- low quality
- signal to noise ratio
- multi modal
- wavelet transform
- ground truth
- probabilistic model