Gottfried Haider | 25 Jun 21:42 2016
Picon

Identifying speech synthesis software

Hello,

I came across this mailing list when looking for communities concerned
with speech synthesis - I hope my request is not too off-topic.

Could anyone help me identifying the software used in this video:
https://www.youtube.com/watch?v=W-luS-LQiDk

Kind regards
Gottfried

Apurva Joshi | 21 Dec 08:37 2015
Picon

how to write our own phrasing rule

guyzz i want to write my own phrase break rule depending on punctuation(" , .)
and after phrase break i want to give different duration pause depending upon punctuation and then want to apply intonation accents to that detected phrase ...so plz tell me how to achieve this prosody modification...

guyzz actually this is last part of my project ...plz help me asap

plz do reply on apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
Apurva Joshi | 18 Dec 10:45 2015
Picon

Fwd: cart tree related query


---------- Forwarded message ----------
From: Apurva Joshi <apurvajoshi1992 <at> gmail.com>
Date: Fri, Dec 18, 2015 at 3:14 PM
Subject: Re: cart tree related query
To: Nageshwara Rao Moova <mnrao2001-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>


hello

i also tried following


(set! simple_phrase_cart_tree    <==given by  phrase.scm
> '
> ((lisp_token_end_punc in ("?" "." ":"))
>   ((BB))
>   ((lisp_token_end_punc in ("'" "\"" "," ";"))
>    ((B))
>    ((n.name is 0)  ;; end of utterance
>     ((BB))
>     ((NB))))))

(set! f2b_phrase_cart_tree              <==given by  phrase.scm
>
> '
>
> ((gpos is punc)
>
>  (((1 0.00238095) (3 0) (4 0.000619) B))         < == here i changed value of B
>
>  (((4 0.00238095) (3 0) (1 0.707619) NB))))


(Parameter.set 'Phrase_Method simple_phrase_cart_tree)

(SayText "कम्प्यूटर, का, प्रयोग,
�रें ")

in above i changed values of B ...but m not getting lot change...i didnt find any change...

plz help




On Fri, Dec 18, 2015 at 3:11 PM, Apurva Joshi <apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
hello Nageshwara sir,

i want to define my own rule for phrase breaking and also want to add different timing silence as per phrase break rule depending on punctuation ...

so how can i write that using lisp function

for eg ...my phrase rule name is mytree

 lets my rule are ....
(set! mytree ' ((R:Token.parent.punc in ("?" "." ":")) ((BB)) ((R:Token.parent.punc in ("'" "\"" "," ";")) ((B)) ((n.name is 0) ((BB)) ((NB))))))

then my default 'Phrase_Method is prob_models

now lets say i set my Phrase_Method to my above "mytree"
ie (Parameter.set 'Phrase_Method mytree)

now when i run (SayText "hello world")

then i get following error ..

SIOD ERROR: not a symbol or string : ((R:Token.parent.punc in ("?" "." ":")) ((BB)) ((R:Token.parent.punc in ("'" """ "," ";")) ((B)) ((n.name is 0) ((BB)) ((NB)))))


So ...plz tell me how to write our own function or module...and here BB means big break...B means simple break

but what is its exactly value(eg 1 sec break ??) ...

and how to change it??

 


Apurva Joshi | 18 Dec 10:41 2015
Picon

Fwd: cart tree related query


---------- Forwarded message ----------
From: Apurva Joshi <apurvajoshi1992 <at> gmail.com>
Date: Fri, Dec 18, 2015 at 3:11 PM
Subject: cart tree related query
To: Nageshwara Rao Moova <mnrao2001-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>


hello all

i want to define my own rule for phrase breaking and also want to add different timing silence as per phrase break rule depending on punctuation ...

so how can i write that using lisp function

for eg ...my phrase rule name is mytree

 lets my rule are ....
(set! mytree ' ((R:Token.parent.punc in ("?" "." ":")) ((BB)) ((R:Token.parent.punc in ("'" "\"" "," ";")) ((B)) ((n.name is 0) ((BB)) ((NB))))))

then my default 'Phrase_Method is prob_models

now lets say i set my Phrase_Method to my above "mytree"
ie (Parameter.set 'Phrase_Method mytree)

now when i run (SayText "hello world")

then i get following error ..

SIOD ERROR: not a symbol or string : ((R:Token.parent.punc in ("?" "." ":")) ((BB)) ((R:Token.parent.punc in ("'" """ "," ";")) ((B)) ((n.name is 0) ((BB)) ((NB)))))


So ...plz tell me how to write our own function or module...and here BB means big break...B means simple break

but what is its exactly value(eg 1 sec break ??) ...

and how to change it??

 

Apurva Joshi | 18 Dec 05:50 2015
Picon

voice modification query

hello everyone

i modified my duration, speed , tempo of my speech correctly that whatever speech i get at the op is totally intelligible ...

even if speech is understandable ... its somewhat sounds like robotic voice ...so i want to make it like human voice..or i want to make voice smooth

so my question is whether it is possible to make voice like human?
but what i think is ..as voice comes as per recording voice..so is it possible to make it smooth or to change complete voice which will be near to normal human voice??

plz help

and plz do reply on apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
Apurva Joshi | 17 Dec 10:32 2015
Picon

voice modification

hello all,

i got different parameter such as duration_stretch ,

(set! duffint_params '((start 90) (end 100)))..

duration changes - speed of speaking

intonation --- just rise or decrease level of voice

my sentence is पुस्तक दस दिनों तक रखें

here i want some gap between two words ...eg i want some gap between पुस्तक and दस ..


so plz help me ...and do reply on apurvajoshi1992 <at> gmail.com


its urgent guyzz ..plz help

Tomoki Toda | 19 Nov 01:04 2015
Picon

Call for Participation: Voice Conversion Challenge 2016

#We apologize for multiple postings.

Dear all,

We are glad to invite you to participate in the first
"Voice Conversion Challenge" that we are organizing soon.

The purpose of this challenge is to compare different systems
and approaches on the task of converting five source speakers'
voices to five different target speakers' voices (females and
males included) from a fixed corpora (both training and
evaluation sets will be provided). In total, 25 conversion
cases will be evaluated in terms of objective audio quality
measures and perceived naturalness and similarity (listening tests).

In this first challenge we focus on 16 kHz speech and
signal-to-signal conversion strategies. No transcription will be
provided and the use of phonetic information is NOT allowed.
However, participants are free of using additional data (for
training purposes) different from the database from which
the speakers corpora will be provided.

The intended schedule is as follows:
- November 23rd: release of training data
  (and notification of participation deadline)
- January 4th: release of evaluation data
- January 11th: deadline to submit the converted audio
-  February 25th-29th: notification of results

There is no fee for registration, COLIPS
(http://www.colips.org/ <http://www.colips.org/>) has generously
provided financial support to this challenge. We invite however
the participants to notify your intention to participate by
*** November 23rd ***. The challenge guidelines that
the participants should strictly follow will be provided in detail
when releasing the training data.

Please contact us here for questions and notification of
participation: vcc2016@...

Organizers:

Lingui Chen (University of Science and Technology of China)
Daisuke Saito (Tokyo University)
Tomoki Toda (Nagoya University)
Fernando Villavicencio (National Institute of Informatics)
Zhizheng Wu (University of Edinburgh)
Junichi Yamagishi (National Institute of Informatics)

Looking forward to hear from you.

Regards,
--

-- 
Voice Conversion Challenge 2016 Organizers
E-mail: vcc2016@...

Apurva Joshi | 8 Oct 14:58 2015
Picon

command for running voice in interactive mode with voice selection

hello guyzz....we know that festival operates in two modes...command mode and interactive mode...

now i want to synthesize in interactive mode

few of commands are:

text2wave .....
echo "hello"|festival --tts

above are the command that we run on terminal ie without going into festival....
but above command run with default voice in festival...ie kal_diphone...in my case...

i have designed hindi voice...

and on terminal i want to fst select hindi voice and then give file as ip to festival...

so is such kind of command available..that will let us to first select voice and then give file as ip to it????

OR

how to set particular voice as default voice??  <==for this question i have tried solutions given by many sites, but m nt getting exactly what to do...?


i want this bcoz ...i want to write shell script in linux which will do this above operation...i have written shell script which successfully runs..and that also enters into festival...but once scrpit reaches to festival then it cant execute command in festival command mode....

eg.... my shell script

#!/bin/bash
cd
$FESTIVALDIR/bin/festival
(voice.list)
(voice_hindi)
(tts "get.txt" nil)

my shell script execute command 1,2,3

after 3 i enter into festival command mode, but after that script cant execute commands 4,5,6....

so plz tell me is following kind of command available?

e.g. echo "नमस्ते " | festival voice_hindi --tts <====  this kind of command which will let us to select voice..

plzzz help....
and plz reply on same id :  apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org



Apurva Joshi | 29 Sep 14:43 2015
Picon

transliteration related

hello guyzzz....
i have one question

whatever the hindi tts i have designed, that takes hindi sentence as input in following form

Aap kaisee hain?  ===> then festival converts that to output speech(hindi speech output.)

now instead of giving input as
Aap kaisee hain?  == > i want to give it as आप कैसीं हैं?....

actually i have designed Machine translation using moses...which will give me output as आप कैसीं हैं?

now i want to give this आप कैसीं हैं? as input to my hindi tts...but as i told u above, my hindi tts accepts input in " Aap kaisee hain? "   <== this form

so i want to transliterate hindi text to english....

my idea is....

1) convert आप कैसीं हैं?  ===> to its corresponding unicode using text to unicode conversion program in cpp...(this unicode will be for hindi letters eg unicode of आ is u0906)

2) map this hindi letters to corresponding english unicode (eg आ  ==>  aa  ....unicode of aa is u0041u0041)

3) after getting english unicode in above step ===> convert this english unicode to text ( आप == aap) using unicode to text in cpp


i have not started unicode to text and vice - versa pgming...but my problem is how to do 2nd step...ie mapping of unicode of hindi to that of english...

plz help...

if anyone of you have any other idea of doing transliteration, then plz reply...i really need help

plz reply on same id

Apurva Joshi | 29 Sep 13:35 2015
Picon

Fwd: problem in adding newly created voice into festival/lib/voices folder


---------- Forwarded message ----------
From: Apurva Joshi <apurvajoshi1992 <at> gmail.com>
Date: Tue, Sep 29, 2015 at 5:04 PM
Subject: Re: problem in adding newly created voice into festival/lib/voices folder
To: Malle ram reddy <ramreddy.malle-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>


Hi Malle ,
thank you for ur reply, i already tried  " ./bin/do_clustergen festvox_dist" this command...but after running this command , its showing error that --

tar: festival/lib/voices/hin/iiit_hin_sukh_cg/README: Cannot stat: No such file or directory
tar: festival/lib/voices/hin/iiit_hin_sukh_cg/COPYING: Cannot stat: No such file or directory
tar: Exiting with failure status due to previous errors

even after this , tar file with name "festvox_iiit_hin_sukh_cg.tar.gz" got created in my hindi voice folder, then i added hindi voice to my festival voice path in following way,

cd $FESTIVALDIR
cd $FESTIVALDIR/lib/voices
mkdir hindi
cd hindi
cp ~/iiit_hin_sukh/festvox_iiit_hin_sukh_cg.tar.gz .
tar xvf festvox_iiit_hin_sukh_cg.tar.gz


after this, " iiit_hin_sukh_cg " , folder get copied in festival/lib/voices/hindi/
then when  i enter festival, and  after entering , i typed command -

festival> (voice.list)
(iiit_hin_sukh_cg
 cmu_us_ksp_arctic_clunits
 kal_diphone
 cmu_us_ksp_arctic)

now when i load hindi voice

festival> (voice_iiit_hin_sukh_cg)
SIOD ERROR: could not open file festvox/iiit_hin_sukh_lex.lookup.scm
closing a file left open: /home/techmahindra/coep_tts/festival/lib/voices/hindi/iiit_hin_sukh_cg/festvox/iiit_hin_sukh_lexicon.scm
closing a file left open: /home/techmahindra/coep_tts/festival/lib/voices/hindi/iiit_hin_sukh_cg/festvox/iiit_hin_sukh_cg.scm


so not getting why that " iiit_hin_sukh_lex.lookup.scm " is not opening , well lexicon.scm calls this file...while building voice, this file has worked , but dunno why at this time its showing error...







On Tue, Sep 29, 2015 at 2:43 PM, Malle ram reddy <ramreddy.malle-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
Hi Apurva,

Just you have to make the voice models as a tar package and copy them into festival/lib folder for automatic calling of your voice models.

"./bin/do_clustergen festvox_dist" i think this will solve your problem, if not do it manually



On Tue, Sep 29, 2015 at 11:30 AM, Apurva Joshi <apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
hello guyzz..,plz help
i have built hindi voice named "iiit_hin_sukh_cg" using procedure given in bsv.pdf to build new voice using clustergen synthesizer.

bsv.pdf/building clustergen synthesizer_page = http://festvox.org/bsv/c3174.html


my built hindi voice is in folder - ~/iiit_hin_sukh
$FESTIVALDIR          -      /home/techmahindra/coep_tts/festival
$ESTDIR                   -      /home/techmahindra/coep_tts/speech_tools
$FESTVOXDIR          -     /home/techmahindra/coep_tts/festvox  

now to run my hindi voice ... i use following command as per given by above link

techmahindra <at> techmahindra:~/iiit_hin_sukh$ festival festvox/iiit_hin_sukh_cg.scm

after that in festival command mode i write following command to load this hindi voice

festival> (voice_iiit_hin_sukh_cg)

after this , voice get loaded perfectly and its speaks well also...now my problem is

the above voice which i generated = iiit_hin_sukh_cg is not available in my festival voice path i.e /home/techmahindra/coep_tts/festival/lib/voices

so every time , when i want to run this hindi voice, i need to go to my hindi folder and then in that folder i run  festival festvox/iiit_hin_sukh_cg.scm this...


and when i want english voice, i simple type $FESTIVALDIR/bin/festival

so , please tell me how can i add my iiit_hin_sukh_cg (hindi voice) into festival voice path ...so that all english , hindi voices i can access with just this - $FESTIVALDIR/bin/festival command....plz i need help asap ....plz rply on same id
apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org




--
Regards
Ram Reddy. M
NERTU,Osmania University


Apurva Joshi | 29 Sep 08:00 2015
Picon

problem in adding newly created voice into festival/lib/voices folder

hello guyzz..,plz help
i have built hindi voice named "iiit_hin_sukh_cg" using procedure given in bsv.pdf to build new voice using clustergen synthesizer.

bsv.pdf/building clustergen synthesizer_page = http://festvox.org/bsv/c3174.html


my built hindi voice is in folder - ~/iiit_hin_sukh
$FESTIVALDIR          -      /home/techmahindra/coep_tts/festival
$ESTDIR                   -      /home/techmahindra/coep_tts/speech_tools
$FESTVOXDIR          -     /home/techmahindra/coep_tts/festvox  

now to run my hindi voice ... i use following command as per given by above link

techmahindra <at> techmahindra:~/iiit_hin_sukh$ festival festvox/iiit_hin_sukh_cg.scm

after that in festival command mode i write following command to load this hindi voice

festival> (voice_iiit_hin_sukh_cg)

after this , voice get loaded perfectly and its speaks well also...now my problem is

the above voice which i generated = iiit_hin_sukh_cg is not available in my festival voice path i.e /home/techmahindra/coep_tts/festival/lib/voices

so every time , when i want to run this hindi voice, i need to go to my hindi folder and then in that folder i run  festival festvox/iiit_hin_sukh_cg.scm this...


and when i want english voice, i simple type $FESTIVALDIR/bin/festival

so , please tell me how can i add my iiit_hin_sukh_cg (hindi voice) into festival voice path ...so that all english , hindi voices i can access with just this - $FESTIVALDIR/bin/festival command....plz i need help asap ....plz rply on same id
apurvajoshi1992-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org


Gmane