[seg.word for seg in decoder.seg()] return None

[Loc Le Duc]

Hi

After I train Acoustic Model with few sample sentences as per the instruction in the topic:
https://sourceforge.net/p/cmusphinx/discussion/sphinx4/thread/f7fa87f5/

I continuous try to decoder using python. I'm flowing this example:
https://github.com/bambocher/pocketsphinx-python

The example run exactly. But when i try to decoder my projects, the results return None.

This is my code:

from os import environ, path

from pocketsphinx.pocketsphinx import *
from sphinxbase.sphinxbase import *

# Create a decoder with certain model
config = Decoder.default_config()
config.set_string('-hmm', 'model_parameters/dkmt.ci_cont')
config.set_string('-lm', 'etc/dkmt.lm.DMP')
config.set_string('-dict', 'etc/dkmt.dic')
decoder = Decoder(config)

# Decode streaming data.
decoder.start_utt()
stream = open('test.wav', 'rb')
while True:
  buf = stream.read(1024)
  if buf:
    decoder.process_raw(buf, False, False)
  else:
    break
decoder.end_utt()
print ('Best hypothesis segments: ', [seg.word for seg in decoder.seg()])

Resuls in terminal:

('Best hypothesis segments: ', [])

I have attached all results in the terminal in this link:
https://www.dropbox.com/s/ra7o8k7aruxkapt/terminal?dl=0

Can anyone help me to sort out the issue.

Thanks and Regards,
DucLoc.

[Nickolay V. Shmyrev]

You need to add a line

 -cmninit 69,3,-4

into the file dkmt.ci_cont/feat.params

[Loc Le Duc]

Thank you very much, Nickolay V. Shmyrev.
Can you help me to explain the number in the line?

-cmninit 69,3,-4

And one more question, do you have any document to learn Sphinx? I want to improve the accuracy and adapt with other people who will use my system.
Thank you very much.

[Loc Le Duc]

Oh, I had found it in the terminal.
But do you suggest any document to learn Sphinx? I want to learn about the issue and develop it for my system.
Thank you very much.

[Nickolay V. Shmyrev]

Our documentation is here

http://cmusphinx.sourceforge.net/wiki