Section Introduction. Recent Approaches to Modeling the Glottal Source for TTS
One of the most important discoveries in speech science was the finding that voiced speech, which accounts for the major part of most utterances, could be fairly well modeled by a pulsive source function serving as excitation for an all-pole transfer function. The source function represents airflow through the glottis, and the transfer function captures the filtering effect of the vocal tract.
KeywordsCoherence Eter Tral Acoustics Santen
Unable to display preview. Download preview PDF.
- [Atal and Remde] B. S. Atal and J. R. Remde. A new model of LPC excitation for producing natural-sounding speech at low bit rates. Proceedings International Conference on Acoustics, Speech, and Signal Processing. ICASSP-82, 614–617, 1982.Google Scholar
- [Macchi et al] M. J. Macchi, M J. Altom, D. Kahn, S. Singhal, and M. Spiegel. Intelligibility as a function of speech coding method for template-based speech synthesis, Eurospeech. In Proceedings Eurospeech’ 93, 893–896, 1993.Google Scholar