| US 7,565,292 B2 | ||
| Quantitative model for formant dynamics and contextually assimilated reduction in fluent speech | ||
| Li Deng, Sammamish, Wash. (US); Alejandro Acero, Bellevue, Wash. (US); and Dong Yu, Kirkland, Wash. (US) | ||
| Assigned to Micriosoft Corporation, Remond, Wash. (US) | ||
| Filed on Sep. 17, 2004, as Appl. No. 10/944,262. | ||
| Prior Publication US 2006/0074676 A1, Apr. 06, 2006 | ||
| This patent is subject to a terminal disclaimer. | ||
| Int. Cl. G10L 13/00 (2006.01) | ||
| U.S. Cl. 704—260 | 18 Claims |
| 1. A method comprising:
receiving a text;
parsing the text into a sequence of phonetic units;
identifying a sequence of target values for a formant based on the sequence of phonetic units;
applying the sequence of target values to a finite impulse response filter to produce a sequence of formant values; and
using the sequence of formant values to generate synthesized speech.
|