| Paper: | SP-P9.6 | ||
| Session: | Topics in Speech Synthesis | ||
| Time: | Wednesday, May 19, 15:30 - 17:30 | ||
| Presentation: | Poster | ||
| Topic: | Speech Processing: Speech Synthesis (including TTS) | ||
| Title: | PROBABILITY BASED PROSODY MODEL FOR UNIT SELECTION | ||
| Authors: | Xi Jun Ma; IBM China Research Laboratory | ||
| Wei Zhang; IBM China Research Laboratory | |||
| Wei Bin Zhu; IBM China Research Laboratory | |||
| Qin Shi; IBM China Research Laboratory | |||
| Ling Jin; IBM China Research Laboratory | |||
| Abstract: | Most modern text-to-speech (TTS) systems are unit selection style. In this kind of systems, the predicted prosody values, such as pitch, duration and energy values for each synthesis unit, are important factors to conduct unit selection. In this paper, a probability based prosody model is presented. In the model, the distribution of prosody values in a given context equivalent cluster is described by Gaussian mixture model (GMM), and the distance between a candidate unit and the context equivalent cluster is defined by probability output of GMM. Then a novel framework for unit selection style TTS systems is derived from the model, and a series of experiments are done on the framework. | ||
| Back | |||
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops