I want to make a project in which my voice command (left, up, etc) get sampled and perform fft, and depending on my voice command it should perform some action.
After Reading above tutorial and code, I understand how to sample and perform fft. and after doing this a got some values(2,0,-4,-1...) .
here I am facing problem how to use these fft value in firing an action.
for example : if I say "left" then arduino get sampled data through adc and perform fft and generate data stream, now using these data stream how I decide that I had spoken "left" or other command or it is a noise...?
I am a newbie so may be it a stupid question,
please help..
for example : if I say "left" then arduino get sampled data through adc and perform fft and generate data stream
But, you can't be bothered to show the code that does this, nor have you confirmed that the data stream is consistently the same each time you say the same thing (which I seriously doubt)
now using these data stream how I decide that I had spoken "left" or other command or it is a noise...?
if(CompareDataStreams(curremtStream, leftStram) == 0)
{
// turn left
}
All that is left for you to do is to define appropriate types for currentStream and leftStream, properly populate leftStream, and srite the function CompareDataStreams(). You should be done in no time.
The only way I can see to do it is to compare the received sound against a library of known possible commands and see whether it was acceptably close to any of them. This sounds like quite a stretch for a little microcontroller to do. You might think about getting something using the EasyVR shield working first before you try to implement your own solution from scratch in software.
