German

Flat
Problems with microphone accuracy
User: falco
Date: 7/30/2009 5:50 am
Views: 8181
Rating: 10

Hello everybody


I have some problems with the accuracy of the German acoustic model.  I tested the sphinx 4 hellodigits and wavfile example. The English model works fine. It was able to recognize nearly 90%.
But the German one gives really bad results. At it’s best it might recognize 10% of the spoken word.  I recorded  16KHz wav files and tested them at the wavfile demo. The results were better but only up to 50% recognition.
The http://www.mediafire.com/?j1l9d0ujmgg test.wav file will be recognize complete correctly.


I tried following files in various combinations:

  • Example files, Voxforge Acoustic model and Sphinx Version from (march 2008) http://www.mediafire.com/?j1l9d0ujmgg  with and without adaption of the “Umlaute” of the dictionary
  • The newest Voxforge acoustic model archive from (march 2009) with and without adaption of the “umlaute” of the dictionary
  • The newest nightly build of sphinx 4 and its examples

I have only used the config file of the http://www.mediafire.com/?j1l9d0ujmgg example and made some adaptations for the microphone support.

What goes wrong here? I use Eclipse with java SE 1.6.0_14 under win XP for testing. I use also two different micros.


It would be nice if you can help me.

Re: Problems with microphone accuracy
User: nsh
Date: 7/30/2009 5:01 pm
Views: 66
Rating: 10

> And made some adaptations for the microphone support.


Probably corrections were not accurate. There can be multiple issues here:

1. sphinx4 doesn't work well on windows with microphone (known, should be fixed in trunk).

2. The wave file you were trying to recognize had wrong sample rate.

I suggest you to upload your modification so we could look and try ourselves.

Re: Problems with microphone accuracy
User: falco
Date: 7/31/2009 2:59 am
Views: 175
Rating: 9

The sample rate of my wave file is the same like in the test.wav.

If you can give me your configuration file for microphone testing and a recorded wav file then it will be nice, too.

I have only copy the live frontend part of the Hallodigits example to test microphone speech.

Re: Problems with microphone accuracy
User: falco
Date: 8/24/2009 2:00 am
Views: 58
Rating: 7

Please check the accuracy of my demo because I could find the problem of these bad results.

http://www.mediafire.com/?zhli0xttixj

Re: Problems with microphone accuracy
User: nsh
Date: 8/24/2009 6:02 pm
Views: 172
Rating: 9

Please use sphinx4-1.0beta3, it was released recently. In theory it should solve microphone accuracy issues.

 

Re: Problems with microphone accuracy
User: falco
Date: 8/31/2009 3:00 am
Views: 118
Rating: 8

Thank you for replay, but the problem isn't solved.


http://www.mediafire.com/?ya1ysxjmjxx
Here you can download my Sphinx4 Beta3 project. Please, look into and show me my mistake.

Re: Problems with microphone accuracy
User: nsh
Date: 9/6/2009 5:28 pm
Views: 67
Rating: 8

Hi Falco


Sorry for delay, I finally looked into this. I didn't check microphone input but only a wavfile part.


So far my thoughs are:

0) Configuration looks ok.


1) Accuracy is about 40% and it's ok for this type of audio.

2) Audio is not a raw recordings. Recordings with N in the beginning miss the spectrum above 5 kHz. I suppose they were upsampled from 11.25 kHz recordings. That's a bad thing because it doesn't give you a spectrum information you've already lost. You need to train 8kHz model that will use only frequences up to 3500 Hz.


3) The files without N at start have some masking in spectrum. Did you convert them from mp3? It's also critical to use special mp3 decoders to get accuracy. Often it will be plain broken. Again, you need an adaptation/special type of model.

4) DE model obviously is not perfect and needs improvement.

Re: Problems with microphone accuracy
User: nsh
Date: 9/6/2009 5:29 pm
Views: 3772
Rating: 8

Ah, and I forgot. The dictionary we are using is a plain broken. Funf for example needs correction. In short someone should fix the dictionary and rebuild the German model.

 

PreviousNext