Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
im consider my self pretty good at programming for the pic series and i dont have a clue on how to implent such a function.
I would guess some sort of algorythm would be need to analyse the recorded speech and i dont think the pic has enough processing power.
just look at the commercial speech recognition programs, they still have chield diseases
For the dsPIC30 and the dsPIC33 Microchip does have dsPIC30F Speech Recognition Library. Us, the free users cant use it because you have to pay for it as its a 3ed party that made it for Microchip.
What can we do? well, read the info on the microchip web site and maybe you will comeup with your own code...
**broken link removed**
The that microchip sales is only for U.S english only, and only for 100 known words. So, Good luck.
I too haven't worked with speech recognition but as far as simple speech recognition is concerned (as implemented in low end cell phones) the idea is to record the sound in digital form, and then when someone repeats the same words the sound is agian recorded digitally and compared with the orignal recorded data.
Now if the comparision is within a certain limit let us say 80% similar to orignal recorded sound then the system thinks that the orignal word has been repeated.
Differnet techniques can be used to do the comparision. One way is to add all the digital samples for the orignal recorded sound this will give a certain figure. Similarly add the digital samples for the word to be compared and then check the difference between between the two sums. If it falls within 80% of the orignal sum then chances are that the orignal word has been repeated.
Any ways that is just my own idea.