festvox-ked-diphone - American English male speaker "Kurt" for Festival
|License:||MIT and GPL+ and TCL|
American English male speaker ("Kurt") for Festival. This voice provides an American English male voice using a residual excited LPC diphone synthesis method. It uses the CMU Lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree.
|festvox-ked-diphone-0.19990610-18.el6.noarch [3.9 MiB]||
by Ray Strode (2010-01-19):
Resolves: #553278 - Make work