modernmedici / voiceid

Automatically exported from code.google.com/p/voiceid
0 stars 0 forks source link

While adding voice model, flt.3.seg file is created as empty #13

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago
1.I am trying to add a voice model for file s2_new.wav. I give the following 
command:

python scripts /vid -j "C:\Python27\share\voiceid\LIUM_SpkDiarization-4.7.jar" 
-b "C:\Python27\share\voiceid\ubm.gmm" -s "testing1" -g 
"C:\Pytho27\pdtrain\s2_new.wav"

I get the following error:
File C:\Python27\pdtrain\s2_new_flt.3.seg empty

I see that s2_new_flt.3.seg has been created but is 0 in size. I also see other 
intermediator files like .adj.3.seg, .c.gmm, .d.3.gmm, etc created and all 
these are not 0 in size.

Can you please tell me on why flt.3.seg is being created as empty file and 
hence adding voice model is not working. I am attaching the input s2_new.wav. 
s2_new.wav has following properties - (1,2,16000, 17368, 'NONE", 'not 
compressed')

thanks a lot,
Regards,
Priy Dixit

Original issue reported on code.google.com by priydp...@gmail.com on 8 Jun 2013 at 2:53

Attachments:

GoogleCodeExporter commented 9 years ago
Hi,
I tried your file and got the same result. By the way I listened to the wav 
file and I am wondering if you are interested in text dependent speaker 
recognition, in this case Voiceid is not a choice.
The voice in the recording is less than a second of speech and our tool and the 
underling Lium library need to have at least 30secs to make a good voice model 
and at least 5-8secs to recognize a voice. If you are lucky could be less, but 
a sec is definitely insufficient.

Original comment by maurome...@gmail.com on 11 Jun 2013 at 1:00

GoogleCodeExporter commented 9 years ago
Hi, thanks for your reply. I tried with a longer 20 second file and it worked ! 
voice id rocks !

Original comment by priydp...@gmail.com on 14 Jun 2013 at 11:22