hey all,
i'm new to this forum and currently i have been pursuing a project in speech recognition
systems. i am trying to design a text dependent; speaker recognition based system using the DSK
6713 digital signal processor. It involves a voice input (word or phrase) which eventually
drives an electronic lock based upon the text content and also upon the verification of the
identity of the speaker. 

I have been successful in implementing the storage of the voice samples(for approximately 10
secs),decimating the signal and playing it back. But i am experiencing quite a predicament
trying to implement the c codes for the fft and the iir filtering operations. so could anybody
help me out with a code for fft and iir filtering( 4th order) for the voice signal input. I
would be really obliged.   

regards
raghav
r...@gmail.com

(You need to be a member of speech-recognition -- send a blank email to speech-recognition-subscribe@yahoogroups.com )

Raghav-

> i'm new to this forum and currently i have been pursuing a project in speech recognition
systems. i am trying to
> design a text dependent; speaker recognition based system using the DSK 6713 digital
signal processor. It involves a
> voice input (word or phrase) which eventually drives an electronic lock based upon the
text content and also upon the
> verification of the identity of the speaker.
>
> I have been successful in implementing the storage of the voice samples(for approximately
10 secs),decimating the
> signal and playing it back. But i am experiencing quite a predicament trying to implement
the c codes for the fft and
> the iir filtering operations. so could anybody help me out with a code for fft and iir
filtering( 4th order) for the
> voice signal input. I would be really obliged.

Can you explain "decimating the signal and playing it back"?  This implies you're
reducing the sampling rate from the
original input rate.  Are you actually doing that?  Or are you filtering the signal?  If the
latter, what's the
difference with the IIR 4th order filtering you say is giving you trouble?

-Jeff

(You need to be a member of speech-recognition -- send a blank email to speech-recognition-subscribe@yahoogroups.com )

Raghav-

>  i) we hav recorded the signal at sampling frequency 8kHz and den decimated
> it to 4khz...

Then how do you control output sampling rate of DSK 6713 board?  The board is fixed to 8 kHz,
so this means you're
output sampling rate is also 8 kHz.

Please explain how you think decimation is supposed to work in this case.

>  ii) now we need to extract features for which we are using filter band
> method...so need to     implement cheby2 iir filters for three
> bands.(50-500, 500-1000,1000-1500hz)
>
> iii) so actually we need code for fft and iir filtering. "c-code".

IIR filter C code is easily found on the web.

-Jeff
> iv) or if u hav already tried voice recognition wit some other logic plz pass on the code
n the details
> regards
> raghav
> *Raghav B <r...@gmail.com>* wrote:
> On 16/04/07, Jeff Brower <j...@signalogic.com> wrote:
>>
>> Raghav-
>>
>> > i'm new to this forum and currently i have been pursuing a project in
>> speech recognition systems. i am trying to
>> > design a text dependent; speaker recognition based system using the DSK
>> 6713 digital signal processor. It involves a
>> > voice input (word or phrase) which eventually drives an electronic lock
>> based upon the text content and also upon the
>> > verification of the identity of the speaker.
>> >
>> > I have been successful in implementing the storage of the voice
>> samples(for approximately 10 secs),decimating the
>> > signal and playing it back. But i am experiencing quite a predicament
>> trying to implement the c codes for the fft and
>> > the iir filtering operations. so could anybody help me out with a code
>> for fft and iir filtering( 4th order) for the
>> > voice signal input. I would be really obliged.
>>
>> Can you explain "decimating the signal and playing it back"?  This implies
you're reducing the sampling rate from the
>> original input rate.  Are you actually doing that?  Or are you filtering the signal? 
If the latter, what's the
>> difference with the IIR 4th order filtering you say is giving you trouble?
>>
>> -Jeff
>>
>

(You need to be a member of speech-recognition -- send a blank email to speech-recognition-subscribe@yahoogroups.com )