Can't create speaker identification model

[GoogleCodeExporter]

What steps will reproduce the problem?

1. vid -j /Library/Frameworks/Python.framework/Versions/2.7/share/voiceid/ -b 
/Library/Frameworks/Python.framework/Versions/2.7/share/voiceid/ -s GERARD -g 
/Volumes/Media/Voices/Gerard.wav 
Traceback (most recent call last):
  File "/Library/Frameworks/Python.framework/Versions/2.7/bin/vid", line 166, in <module>
    default_db.add_model(file_basename, speaker)
  File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/voiceid/db.py", line 181, in add_model
    fm.build_gmm(basefilename, identifier)
  File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/voiceid/fm.py", line 315, in build_gmm
    diarization_standard(filebasename)
  File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/voiceid/fm.py", line 524, in diarization_standard
    + filebasename)
  File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/voiceid/utils.py", line 79, in start_subprocess
    raise err
OSError: Subprocess <subprocess.Popen object at 0x6eb1d0> closed unexpectedly 
[java -Xmx2048m -jar 
/Library/Frameworks/Python.framework/Versions/2.7/share/voiceid/ 
fr.lium.spkDiarization.system.Diarization --fInputMask=%s.wav 
--sOutputMask=%s.seg --doCEClustering /Volumes/Media/Voices/Gerard_]

Original issue reported on code.google.com by actongor...@gmail.com on 8 Sep 2013 at 5:24

[GoogleCodeExporter]

This command is uncorrect: the -b option set the gmm UBM model path. Probably 
you want set the gmm DATABASE path and the correct option is -d. 

Original comment by michelaf...@gmail.com on 11 Sep 2013 at 11:54

• Changed state: Verified