<html>
<body>
<br>
<font color="#0000FF"><x-tab>
</x-tab>Know that
</font><h1><font size=2 color="#0000FF">(OSLEC)</font>
</h1><font color="#0000FF">pretty well. Wanna test new algorithms but
need wav data. Echo Cancel is also still under research process, VAD is
part of a good echo canceller as it freezes the algo processing in case
of double-talk. Generally a modified Geigel algo or energy based algo are
used.<br><br>
I have new algorithms
and want to test them. Unfortunately, I don't have enough data, there are
out there very expensive databases such as (AURORA...) but don't think
they are really suitable for my tests.<br><br>
Zelda<br><br>
<br><br>
</font>At 01:55 19/02/2008, you wrote:<br>
<blockquote type=cite class=cite cite="">KMTDBOURSE wrote:<br>
> Hi,<br>
><br>
> Thanks a lot for describing the vad algo it seems rather
a basic <br>
> algo and I am surprised it is not more sophisticated. I hear also
from <br>
> speex users (must be same algo?) that the vad algo is very
poor and <br>
> should be improved. That is what I am doing, try to improve vad and
<br>
> echo cancellation algorithms. I would appreciate a link where I can
<br>
> find data files (audio) to download on which I can perform tests. Of
<br>
> course I will feedback the freeswitch community in case of
improvment.<br>
><br>
> best<br>
><br>
> Zelda<br>
There is a pretty good free echo canceller at
<a href="http://www.rowetel.com/" eudora="autourl">www.rowetel.com</a>,
although <br>
FreeSwitch can't directly use it as it is GPL software. It does, <br>
however, mean that free and open workable algorithms for robust EC do
exist.<br><br>
VAD is different. It is still a research topic. You can see this from
<br>
the way the VAD algorithms in things like G.729 have been revised over
<br>
the codec's lifetime, as further research has been done. The algorithms
<br>
G.729 and AMR use are encumbered by patents, otherwise cloning them <br>
would be excellent start. VAD for speech recognition is relatively <br>
straightforward. When you can buffer the audio, and look forwards and
<br>
backwards in time, it become much easier to identify the start and end
<br>
of voice bursts. It becomes hard when you can only look back in a low
<br>
latency scenario.<br><br>
Steve<br><br>
<br>
_______________________________________________<br>
Freeswitch-dev mailing list<br>
Freeswitch-dev@lists.freeswitch.org<br>
<a href="http://lists.freeswitch.org/mailman/listinfo/freeswitch-dev" eudora="autourl">
http://lists.freeswitch.org/mailman/listinfo/freeswitch-dev</a><br>
UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-dev<br>
<a href="http://www.freeswitch.org/" eudora="autourl">
http://www.freeswitch.org</a></blockquote></body>
</html>