Voice recognition in labview

[riah1920]

I am doing a project using voice controlled wheelchair.I am new to labVIEW.I have done the interfacing wifi connection with myRIO and making a dc motor to run using myRIO.I have decided to use Microsoft Speech API to convert speech to text.Is it possible to interface the speech software in LabVIEW.Please reply as soon as possible .

[riah1920]

I am doing a project in Voice Controlled WheelChair using LabVIEW and i have complications in giving the input.I have recorded the commands and given as an input.But the input wav file is not accepted.Please help me with this

[GerdW]

Hi riah,

Is it possible to interface the speech software in LabVIEW.

Using LabVIEW on WindowsPC: probably.

Using LabVIEW-RT/FPGA on your myRIO: no.

[riah1920]

Thankyou Sir.So you are saying I can't use the software incase I am using MyRIO.Please tell me if it is it possible to record all the commands in a .wav file and storing in an array format and then giving a voice input from myRIO to compare with the commands I have stored  to run the dc motor.

[GerdW]

Hi riah,

surely you could store some waveforms in your myRIO and surely you can try to compare them with recorded sound.

But voice recognition is more than just comparing two waveforms - not to mention the noise in all waveforms you want to compare…

[riah1920]

thankyou for the reply.... 

[riah1920]

I have done the comparision of two waveforms. But i am not able to compare an array of waveforms with a single waveform.Please suggest me some solution.

[GerdW]

Hi riah,

so you know how to do "one" operation - but you struggle to do "more than one, but the same" operation?

What about an autoindexing loop???

I still don't think comparing waveforms of recorded sound for being the "very same" will allow you solve your voice recognition problem…

[riah1920]

My project is not voice recognition..... It is only the recognition of commands...... forward,backward,left,right and stop...... Will comparing waveforms will help in this case......

[JÞB]

You could dig under the hood of "LV Speak" it has a cross platform speech recognition module that may give you some pointers.

(Or possibly provide the answer)