[hts-users:01843] Re: Where to start with HTS

[Sacha Krstulovic <sacha@xxxxxxxxxxxxxxxxx>]

IMO a good starting point is to analyze and understand
the HTS demo scripts, e.g.,
 http://hts.sp.nitech.ac.jp/archives/2.1/HTS-demo_CMU-ARCTIC-SLT.tar.bz2
They contain some examples of both training the models and
running hts_engine. Then you can also use the HTK book as
a reference to learn to train and use HMMs with the HTS/HTK modules.

One crucial point to understand is that HTS doesn't come with
a text analysis front-end (the function Festival is used for
in the demos) so for synthesizing any new language you will
have to provide an external front-end. The front-end is the part
which goes from the text to the full-context labels. So you need
HTK/HTS to train the full-context models and hts_engine
(or HMGenS+SPTK) to synthesize speech, but for both you need
your own separate front-end. Using Festival as a front-end is
an option if Festival covers your language or if you are
familiar with programming in LISP/Scheme. Otherwise you
will have to provide your own front-end.

All this requires a fair bit of reading and self-teaching
about HMM synthesis. Many very good publications are available,
e.g., trough the HTS homepage, and again the HTS demos are
a very good example of the whole process. The mailing list
will help if you get stuck with precise technical points,
but it won't replace a well-grounded general understanding
of the method and tools.

HTH;

-*- Sacha K. -*-


Mohamed Abdou wrote:
> Hi all,
> I wonder where to start to try running HTS to synthesis speech, and do I 
> need HTK if I am dealing with hts_engine or only Festival. Another thing 
> is there a demo same like chpater 3 in htk book showing an exmaple 
> running hts_engine knowing that I am working on a new language and 
> prepared my speech database.
>  Thanks in Advnce
> M. Abdou
>
> ------------------------------------------------------------------------
> Discover the new Windows Vista Learn more! 
> <http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please visit http://www.messagelabs.com/email
> ______________________________________________________________________

--
Dr.Sacha Krstulovic - Research Engineer
Toshiba Research Europe Limited
Cambridge Research Laboratory
Speech Technology Group
208 Science Park, Milton Road
Cambridge CB4 0GZ - United Kingdom
Tel:    +44 1223 436 978
Fax:    +44 1223 436 909
E-mail: sacha@xxxxxxxxxxxxxxxxx



______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email 
______________________________________________________________________
begin:vcard
fn:Sacha Krstulovic
n:Krstulovic;Sacha
org:Toshiba Research Europe Limited - Cambridge Research Laboratory;Speech Technology Group
adr:;;260 Science Park, Milton Road;Cambridge;;CB4 0WE;United Kingdom
email;internet:sacha@xxxxxxxxxxxxxxxxx
title:Research Engineer
tel;work:+44 1223 436 978
tel;fax:+44 1223 436 909
x-mozilla-html:FALSE
url:http://www.toshiba-europe.com/research/crl/
version:2.1
end:vcard