[hts-users:00673] Re: Speaker Adaptation

["Heiga ZEN (Byung Ha CHUN)" <zen@xxxxxxxxxxxxxxx>]

Hi,

Tamer Fares wrote (2007/05/16 16:47):

> 1. prepare the utterances for the adaptation
> 2. using the HHEd to build a regression class trees
> 3. perform static adaptation using the HERest to perform global
> adaptation then use this global adaptation as an input transformation to
> transform the model set
> 4. how can i use then the HMGens to generate the new parameters to be synthesis ?

You can load this global adaptation transforms as an input transforms of HMGenS using the -J option.
To turn on input transform, -a option is required.
Of course baseclass & regression tree mmfs should be loaded. 

Regards,

Heiga ZEN (Byung Ha CHUN)

--
------------------------------------------------
Heiga ZEN     (in Japanese pronunciation)
Byung Ha CHUN (in Korean pronunciation)

Department of Computer Science and Engineering
Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya 466-8555 Japan

http://www.sp.nitech.ac.jp/~zen
------------------------------------------------