Analyzing the tonal content of someone's voice and recognizing different people voices

[Jessica17]

How does a VI look like that recognizes particular sound and raises an alarm? I also need to be able to analyze the tonal content of someone's voice so that I can differentiate between several people.. As I am beginner, is it possible to do this in a simple way in labview or maybe I should use another program?

[Bob_Schor]

What do you know about Signal Theory, particularly as it pertains to the Voice Recognition problem you ask?  Once you know what you want to do (including knowing about sampling rates, necessary precision for processing voice information, other pertinent parameters that are peculiar to your task), you can begin to consider how to accomplish your task.  Is LabVIEW appropriate?  Probably, but you'll also have to think about how you acquire the data (using LabVIEW?  something else?) and what processing requirements you have.

If you are new to LabVIEW, this may be too much to handle as a first Project.  If you are really interested in pursuing it, find yourself a LabVIEW Mentor who likes to teach/train mentees and apprentice yourself for a few weeks.  Be prepared to put in some effort.  Having some math and some computing experience will be helpful ...

Bob Schor

[GerdW]

Hi Jessica,

Monday you asked for a simple LUT, now you want to tackle voice recognition?

You really are a fast learner - atleast I guess so from your questions…

Please take the LabVIEW beginner courses before jumping into deep water!

is it possible to do this in a simple way in labview or maybe I should use another program?

Using LabVIEW it will be as simple as in other programming languages. As Bob said: you need to know the theoretical background of your task…

[Jessica17]

I do not have any specific parameters, I just want to make a simple version without many complications that can differentiate the sound of hitting the table with several different objects for example. Just the basics

[GerdW]

Hi Jessica,

differentiate the sound of hitting the table with several different objects

This sounds like a very different task, I wouldn't call it "voice recognition" anymore…

Use a accelerometer attached to your table. Get the readings at sufficient high sample rate. Analyze the signal patterns.

The first two steps are easy with LabVIEW (and the right DAQ hardware). THe last one involves some knowledge on signal analysis - as has been suggested several times before…