Here are some links

wiki actually does a good job of covering it, may want to start there.
http://en.wikipedia.org/wiki/Speech_recognition
http://en.wikipedia.org/wiki/Hidden_Markov_model

http://www.cs.cmu.edu/~rsingh/homepage/papers/zsynchicassp.pdf
http://www.circuitcellar.com/pastissues/articles/Stewart91/text.htm
http://www.edn.com/article/505375-Design_Ideas.php#add
Scroll down for discrete recognition IC to interface with PIC

 

Good night to eastern europe then

thanks! Guess it getting late in da States as well so gudnit all...

 

thanks! Guess it getting late in da States as well so gudnit all...

In states they have nice evening, here in middle europe it is 2:10am so I am going to sleep as well.

 

If you need to store a lot of 10 bit numbers, I came up with a good idea (I think) here:
forum.allaboutcircuits.com/showthread.php?t=56592

 

I hope you aren't going to try to match an analog waveform to another to see if the voice commands match.

You'll need to make a "signature" that can be processed and matched.

you can read my mind...

 

Here is one with a dsPIC using multiple commands

I just seen the video...WOW i like that! i got to do my own...
and if i really acheive this...i'll be the happiest man in the world for 'one week'

 

If you need to store a lot of 10 bit numbers, I came up with a good idea (I think) here:
forum.allaboutcircuits.com/showthread.php?t=56592

i'll definately have a look at your code very soon...

Thx!

 

OK guyz!

i think i need to take one step backward...i anticipated my thing by asking about sampling and RAM storage...i'll get back to it later.

BTW, ima order on monday the following PICs: PIC16F877 (5MIPS, 256byte RAM), PIC18F2455 (12 MIPS, 2Kb RAM) and PIC18F25K80 (16MIPS, 3Kb RAM) i will be using these for my future projects before moving on to PIC32...

But as i have PIC16F690 at the moment, i will work with it while waiting for the above PICs...to implement my SRS...

i'll post a schematics for the hardwad part of the design...but not too interrested in the hardward as the software part is the core of this design...

Also im planning to achieve this by doing this one step at a time...i always tend to try to understand errthing at the same time (not good)

And the first three steps that im going tackle are:
1. detection (capturing) of speech signal;
2. sampling speech signal with ADC + storing of the sampling in RAM for further processing;
3. creating a fingerprint of sample word (i think this is what thatoneguy refered to as signature )

After the above three steps ima work on the second part of the design which will be basically creating a model for the word to be matched with...i'll explain later what i mean.

Alright, i been thinking about implementing 'voice/speech detection' as i think this is the starting point of the system BUT i failed to come up with a good idea or algorithm with it...i think it a bit advanced for me...for now

So i been thinking of the following scenario:
- Adding a pushbutton to the system to allow ADC to process the speech analog signal! Good idea right!?
- Basically the user will have to 'press and hold down' a pushbutton, speak a wordthen the ADC will sample the signal and as soon as the button is released the sampling should stop.
do you think it a good idea?? or maybe after a certain time 'X' if the button is still presseddown the sampling MUST stop! is it better??

But im thinking:
-there should be a 'delay' between the 'press and hold' of the pushbutton and the start of the sampling.
if you think it necessary, how long should be this delay?? 0.5 second? or ??

-secondly these samples must be save in the RAM for further processing, but then PIC16F690 has only 256 bytes of RAM...so how many samples can be saved in that 256 bytes RAM? coz initially i was thinking of storing 2000 samples (which is half a second) with ADC sampled at 4Khz considering humain voice frequency range:~150Hz to ~2000Hz (highest frequency)....

Now given my small 256bytes RAM, by how much do you think i can reduce the number of samples to be stored in the RAM??

Hey i think i can shut up for now...

Thanks all for your comments

 

Now this is interesting. I typed "Speech Recognition" into Microchip's site search engine, and first hit back was for the "dsPIC30F Speech Recognition Library."

Summary

The dsPIC30F Speech Recognition Library provides voice control of embedded applications that require an alternative user interface. With a vocabulary of up to 100 words, the Speech Recognition Library allows users to control their application vocally. The Speech Recognition Library is an ideal front end for hands-free products such as modern appliances, security panels and cellphones. The Speech Recognition Library has very modest memory and processing requirements and is targeted for the dsPIC30F5011, dsPIC30F5013, dsPIC30F6012 and dsPIC30F6014 processors.

Links to the users guide and lib download are right here.

 

Now this is interesting. I typed "Speech Recognition" into Microchip's site search engine, and first hit back was for the "dsPIC30F Speech Recognition Library."

Summary

The dsPIC30F Speech Recognition Library provides voice control of embedded applications that require an alternative user interface. With a vocabulary of up to 100 words, the Speech Recognition Library allows users to control their application vocally. The Speech Recognition Library is an ideal front end for hands-free products such as modern appliances, security panels and cellphones. The Speech Recognition Library has very modest memory and processing requirements and is targeted for the dsPIC30F5011, dsPIC30F5013, dsPIC30F6012 and dsPIC30F6014 processors.

Links to the users guide and lib download are right here.

Thanks ErnieM! i was a bit anxious as i was not getting any comment...
But i think i said that before that i dont want to use a dsPIC as i will have to learn again how to program that type and im not even excellent or very good with the normal PICs

I am totally aware of those dsPic for voice recognition...while doing some research i saw that...

I think i explained how i would like to do it...i have explained the first steps...and this is what i am interrested in for now!!! the rest will follow afterwards...

i want to do this the hard way in order to learn more about PIC! for instance i never used ADC before but now i am etc...

PLEASE guyz help me achieve my three first steps...and forget about this speech thang!! Right now im interrested in activating the sampling using a pushbutton and storing the samples in the RAM. that's all!!!

let's forget about speech recognition...

regards, Eric