[hts-users:02866] Question about duration model structure

["ZhuFengyun" <zhufengyun@xxxxxxxxxx>]

Dear HTS developers:

 

As Dr. ZEN said in http://hts.sp.nitech.ac.jp/hts-users/spool/2007/msg00680.html

 

"In the previous versions we used a multi-variate Gaussian PDF to represent state duration PDFs.
However, since version 2.0.1 we have changed it to multi-stream structure."

 

I'm confused about why are you making the decision that using multi-stream structure instead of multi-variate Gaussian PDF?

Can someone make a brief explanation about the advantages of using a multi-stream structure?

 

Thanks in advance.

 

Best regards,

Zhu Fengyun

------------------------------------------------

Speech and Hearing Research Center

Department of Intelligent Science and Technology

School of Infomation Science and Technology

Peking University

Beijing, China