[hts-users:04490] Re: get bad voice by run HTS-2.3.1

thank you for the reply.

what i was trying to trace out is why HTS is taking too long in voice adaptation

I installed HTS-2.3beta_for_HTK-3.4.1 and using hts api 1.10

up to SI modeling it was fine . but SAT is not completed yet .

Earlier the with the same data it took only 3days 18 hours for 5000 (1000 per speaker) sentences .

but this time after 6days it is still running .

i am working on ubuntu 16.04 LTS

On Wed, Feb 15, 2017 at 1:44 PM, Sébastien Le Maguer <slemaguer@xxxxxxxxxxxxxxxxxxxx> wrote:

Hello,

I'm gonna answer to the 2 mails in on shot :D

@Niya: if you try to know if you are using version 2.3.1 or 2.3, the only difference is on the demo process (the file http://hts.sp.nitech.ac.jp/archives/2.3.1/HTS-demo_CMU-ARCTIC-SLT.tar.bz2 ) which is introducing DNN. If you actually want to check the version of the binaries of HTS there is not direct I know. However, you can take the changelog file ( http://hts.sp.nitech.ac.jp/archives/2.3/HTS_Document.pdf ) and compare the options of binary you have with the new introduced for each version.

@IMUCS_liurui: depends what you defined by no better but, globally, I think that the corpus is actually too small to see the difference. This kind of algorithm needs a lot of data.

Kind regards,
Sébastien

On Wed, Feb 15 2017 (06:06), Niya N <niya121212@xxxxxxxxx> wrote:

> how can i get hts 2.3.1 package
>
> how to check the version of hts currently installed
>
> On Sun, Feb 5, 2017 at 1:05 PM, IMUCS_liurui <liurui_imu@xxxxxxx> wrote:
>
>> I used the same data to run 2.3 and 2.3.1, but found the results of
>> 2.3.1（dnn）is no better than 2.3
>>
>> excuse me, can you tell me the reason?
>> Thank everyone.:)
>>
>>
>>
>>

--
Dr. Sébastien Le Maguer
Postdoctorate researcher

Saarland University
Campus C7.4 - room 2.03
D-66123 Saarbrücken
Germany

phone : +49-681-302-70030
Mail: slemaguer@xxxxxxxxxxxxxxxxxxxx
website : http://www.coli.uni-saarland.de/~slemaguer/