[hts-users:01382] Re: questions about Tree-Based State Tying

[paminy <paminy@xxxxxxx>]

Yes,I've made a big mistake for modeling  hmms on frames rather than on phonemes. I'm sorry for this 
ridiculous mistake for my carelessness. 
Thank you for pointing out the mistake. 

 

在2008-05-06，"Simon King" <Simon.King@xxxxxxxx> 写道：

paminy wrote:
> Thanks for your answer, Simon.
> I still did not understand the "same" pdf.
> Why they have the same PDF ?
> As there are three or five different states and also the pdfs in one 
> frame ,I do not understand why they have the same pdf whith in so many 
> different frames?
>
> If using the same PDF,how dose the pdf's mean and variance caculate?
> Dose it means the pdf's mean is the average of all pdfs?

I think you have some rather fundamental misunderstandings about HMMs. 
You need to read some introductory texts first. Then, look at the HTK 
manual sections on parameter sharing.

Simon

---

网上订比萨天天有奖