[Asterisk-Dev] Voice detection

[Brad Waite]

Chris,

Have you stumbled across any VOX code that would be simpler to introduce to * 
than Sphinx?

Brad Waite

Chris Albertson wrote:

> The "magic words" to use for Google are "VOX DSP Algorithm"
> "VOX" means "Voice Operated Switch" and is very common in
> simplex radio.  It is a kind of automatic "push to talk switch"
> 
> Yes Spinx would have something like this some place in it but
> then so would any teliphony codec that did silence surpression.
> 
> The basic idea is a compute the power in the audio, apply a _very_
> low (on order of about 1Hz) pass filter and threshold the result.
> You may want two thresholds one for on one for off to provide
> stability. (Remember the idea of a dead band in servo systems?)
> Filter responce and the threshholds should be user adjustable
> 
> Yes, I've been looking a spinx for some time now.  I think it is
> "ready for prime time" for simple applcations like voice dialing
> a phone number or yes/no to an IVR system.  
> 
> --- Brad Waite <brad at wcubed.net> wrote:
> 
>>Steven Critchfield wrote:
>>
>>
>>>On Thu, 2003-10-02 at 21:36, Brad Waite wrote:
>>>
>>>
>>>>Does anyone know if there's public voice detection algorithms
>>
>>available?  I've 
>>
>>>>scoured the net for the last hour or so, and I can't come up with
>>
>>anything 
>>
>>>>except a few proprietary or embedded solutions.
>>>>
>>>>I know dsp.c uses goertzel algorithms for DTMF detection, but how
>>
>>does one 
>>
>>>>detect voice?
>>>>
>>>>I dunno, maybe detecting voice isn't the way to go.  I want to
>>
>>begin playback of 
>>
>>>>a file after a phone/answering machine has answered.
>>>
>>>
>>>Argghh telemarketing....
>>
>>Argghh voice mail notification...  (among other non-telemarketing
>>things)  :)
>>
>>
>>>shouldn't you detect silence and then start?
>>
>>Possibly, but I would think that between the last ring and the speech
>>there 
>>could be enough silence to return a false positive.
>>
>>As it turns out, it appears that Festival's cousin, Sphinx, might be
>>the ticket. 
>>  Sphinx not only detects voice, but is a full Open Source speech
>>recognition 
>>engine.  In my mind, that opens up worlds of possibilities with *. 
>>Anyone else 
>>up for it?
>>
>>BTW, more on Sphinx can be found at
>>http://fife.speech.cs.cmu.edu/sphinx/.
>>
>>Brad Waite