Material needed for french model creation

[JRMeyer]

>>> elpimous_robot
[December 19, 2017, 4:09pm]

Hi all, I'd like to create a FRENCH Deepspeech model :

Voxforge provides slash ~140 hours, and is not too difficult to clean. slash
the Lium (university of Mons) provides lot of hours, but audio format
'sph' (sphere) must be converted to wav.

If you know location for others audio material place, please answer
here.

Thanks all

[This is an archived TTS discussion thread from discourse.mozilla.org/t/material-needed-for-french-model-creation]

[JRMeyer]

>>> lissyx
[January 5, 2018, 5:44am]

One thing that we could work on would be to extract audio and text from
french parliament sessions. At least I know the debates' transcriptions
are not 100% accurates, but the written speeches from ministers for
examples, should be.

[Archived Post]

[JRMeyer]

>>> elpimous_robot
[January 5, 2018, 11:07am]

Hi Lissyx, slash
Yes, it could be a good way to gain wav's. slash
Now, need a tool to properly cut sentences, ( because we can't estimate
a pause duration as a endligne ('point')

Audiobooks (librivox) are a good way, too

[Archived Post]

[JRMeyer]

>>> yv001
[January 5, 2018, 12:12pm]

Not sure about French specifically but other European countries have
national language institutions that are building and sharing language
corpora including audio corpora.

The both collect data themselves and use sources like transcribed news
from national TV. You can check if that's available for French.

[Archived Post]

[JRMeyer]

>>> lissyx
[January 5, 2018, 12:16pm]

Interesting, do you have any links for that? I've failed finding
anything at EU level
![:frowning:](

[Archived Post]

[JRMeyer]

>>> yv001
[January 5, 2018, 12:51pm]

I was looking for a national language specifically rather then a unified
EU platform but some raw resources (different licenses, different
formats) for several languages can be found here:

list of
corpora

I suppose, It's better than starting from scratch, especially when used
for non-commercial purposes.

[Archived Post]

[JRMeyer]

>>> elpimous_robot
[January 5, 2018, 12:57pm]

Yes, nice idea.

Question : does anyone has an idea for cutting audio/text ? slash
I think that we must limit sentences duration : a long sentences can
cause bad synchronization. slash
Cutting a sentence at a end point is the normal way, but about Vad ?!

Please, add your ideas (or programs link, LOL)

[Archived Post]

[JRMeyer]

>>> lissyx
[January 5, 2018, 1:11pm]

I wanted (but had no time) to try and use Gentle
slash [https://lowerquality.com/gentle/ slash ] for that on some videos from
assemblee-nationale.fr

[Archived Post]

[JRMeyer]

>>> elpimous_robot
[January 5, 2018, 1:14pm]

Nice, Lissyx. slash
I'll have a loot at this opensource program.

[Archived Post]

[JRMeyer]

>>> Gman
[February 9, 2018, 8:31pm]

elpimous,

Are you still looking to create a french model? slash
Is there anything we can do to help?

I've trid to record me on VoxForg but haven't been validated...so I
suppose the project is dead and halas, Mozilla doesn't give yet the
opportuniy ro record voice other than English.

Regards,

G

[Archived Post]

[JRMeyer]

>>> lissyx
[February 9, 2018, 8:24pm]

Common Voice is being localized, as we speak. So it should be possible
to work on other languages as soon as this is completed.
[ slash mikehenrty](

[Archived Post]

[JRMeyer]

>>> Gman
[February 9, 2018, 8:34pm]

looks a good news let's
hope it will be ready soon!

but if I can help any other way I'll be glad to help

[Archived Post]

[JRMeyer]

>>> elpimous_robot
[February 9, 2018, 9:40pm]

hi [ slash Gman](
Thanks for your proposition. slash
As Lissyx said, I'm waiting for common voice localization... slash
See U pretty soon.

[Archived Post]

[JRMeyer]

>>> mark2
[February 14, 2018, 8:34am]

, I tried Gentle for Finnish
audio file. I guess it uses English acoustic and language models because
most of the words are not recognized:

'words': slash [ slash
{ slash
'alignedWord': '', slash
'case': 'success', slash
'end': 6.430000000000001, slash
'endOffset': 12, slash
'phones': slash [ slash
{ slash
'duration': 0.11, slash
'phone': 'oov_S' slash
} slash
slash ], slash
'start': 6.32, slash
'startOffset': 0, slash
'word': 'Oleskelulupa' slash
}, slash
{ slash
'alignedWord': '', slash
'case': 'success', slash
'end': 7.3999999999999995, slash
'endOffset': 17, slash
'phones': slash [ slash
{ slash
'duration': 0.93, slash
'phone': 'oov_S' slash
} slash
slash ], slash
'start': 6.47, slash
'startOffset': 14, slash
'word': 'Jos' slash
}, slash
{ slash
'alignedWord': '', slash
'case': 'success', slash
'end': 10.83, slash
'endOffset': 23, slash
'phones': slash [ slash
{ slash
'duration': 0.3, slash
'phone': 'oov_S' slash
} slash
slash ], slash
'start': 10.53, slash
'startOffset': 18, slash
'word': 'tulet' slash
} slash
...

So, I don't understand how it will help you with French... If you will
find some workaround, tell us!

[Archived Post]

[JRMeyer]

>>> lissyx
[February 15, 2018, 1:18pm]

Well, from my discussion with the people working on that, I think there
was a piece about making a model for french, obviously
![:slight_smile:](

[Archived Post]

[JRMeyer]

>>> mark2
[February 16, 2018, 8:59am]

I found very good tool for audio and text alignment which supports
several languages including French, Finnish and many others - Aeneas,
https://www.readbeyond.it/aeneas/

[Archived Post]

[JRMeyer]

>>> lissyx
[February 16, 2018, 9:09am]

Nice, however, I guess it requires some pre-processing. They document
the input to be
>,
which means someone has already segmented sentences.

I'd be interested in the results it would give on a long text. Would
automatically cutting off sentences using punctuation works?

My target is stuff like that:
http://videos.assemblee-nationale.fr/video.4738358_595b8f797febf.1ere-seance--ouverture-de-la-session-extraordinaire--declaration-de-politique-generale-du-gouvern-4-juillet-2017?timecode=14408000
that's 4h of audio, with subtitles.

[Archived Post]

[JRMeyer]

>>> mark2
[February 16, 2018, 9:27am]

Yes, you should segment at least text (transcription) file somehow. I
use NLTK-package to split raw text on punctuation or alternatively you
can split anyway you want. As the result I get each segment on its own
row in the transcription file. And then Aeneas does the rest!

There might be very tiny error such as the first or last word goes into
wrong segment, but otherwise it performs surprisingly well.

[Archived Post]

[JRMeyer]

>>> lissyx
[February 16, 2018, 9:30am]

Nice ! Would you be able to give it a try on the video I linked above,
if I extract you audio and text links ?

[Archived Post]

[JRMeyer]

>>> mark2
[February 16, 2018, 9:31am]

Sure, what I need is a single audio-file and a single text file that
contains the transcript of that audio.

[next page
→

[Archived Post]