asterisk users - Aug 2010 - Identify remote prompts: Partial audio matching?

Ok, here's the challenge:

I would like to be able to find, match - and then react - upon prompts 
that are presented by the outbound/remote side of a call. Think mobile 
phone and "This user is temporarily unavailable".

Collecting a limited number of known prompt snippets should not be a 
problem, but how would you then detect their presence in a longer 
recording (or live audio stream)?

Recently there was an at least slightly related posting on this list, if 
I recall that correctly, but I have simply not been able to turn this up.

Philipp

P.S.: This is all about audio analysis, not about cause codes.

>From: asterisk-users-bounces at lists.digium.com
[mailto:asterisk-users-bounces at lists.digium.com] On Behalf Of Philipp von
Klitzing
>Subject: [asterisk-users] Identify remote prompts: Partial audio matching?
>Ok, here's the challenge:
>I would like to be able to find, match - and then react - upon prompts 
that are presented by the outbound/remote side of a call. Think mobile 
phone and "This user is temporarily unavailable".
>Collecting a limited number of known prompt snippets should not be a 
problem, but how would you then detect their presence in a longer 
recording (or live audio stream)?
>Recently there was an at least slightly related posting on this list, if 
I recall that correctly, but I have simply not been able to turn this up.
>Philipp
>P.S.: This is all about audio analysis, not about cause codes.
You might be able to record these snippets then pass them through the Vestec
or Lumenvox Speech engine to get what you want.

On Wed, Aug 4, 2010 at 10:12 PM, Philipp von Klitzing
<klitzing at pool.informatik.rwth-aachen.de>
wrote:
> Ok, here's the challenge:
>
> I would like to be able to find, match - and then react - upon prompts
> that are presented by the outbound/remote side of a call. Think mobile
> phone and "This user is temporarily unavailable".
>
> Collecting a limited number of known prompt snippets should not be a
> problem, but how would you then detect their presence in a longer
> recording (or live audio stream)?
>
> Recently there was an at least slightly related posting on this list, if
> I recall that correctly, but I have simply not been able to turn this up.
>
> Philipp
>
> P.S.: This is all about audio analysis, not about cause codes.
Exact match:
http://github.com/Motiejus/SoundPatty

Regards
Motiejus Jak?tys