[Freeswitch-users] Detect speech in the background with pocketsphinx?

[Peter P GMX]

Hello,

is there a way to detect speech by pocketsphinx in the background?

I want to do the following:
I want to playback al longer sound file to a caller. Whenever the caller 
considers, that this part of the file is good, he can say "good", and 
when he reaches a part of the soundfile which he considers as bad, he 
says "bad".

This means for me, that I have to playback the soundfile to the caller 
by the XML dialplan and that I have to catch some events in the 
background through e.g. event socket. During the playing of the 
soundfile I also would catch these events multiple times, the caller 
should be able to say "good" and "bad" many times during the soundfile. 
So this is not just an IVR application where I react on speech and then 
do something new. In this case the playing of the soundfile must 
continue after a word is recognized.

Questions?
- can we achive this with freeswitch?
- how can this in gerneral be done?

Thanks in advance
Peter