LabVIEW speech recognition

[1206]

Hello, I just want to ask help. I have seen the video about the voice recognition in labview. Can we get the VI's that were used in that technology. Thanks. You're immediate reply is very much appreciated.

[smercurio_fc]

---

@1206 wrote:

Hello, I just want to ask help.

---

Well, that's usually why most people come here.

Unless you're @alfa1. But then, he's probably coming for the same reason.

---

 I have seen the video about the voice recognition in labview.

---

What video? Link?

---

Can we get the VI's that were used in that technology. Thanks. You're immediate reply is very much appreciated.

---

As asked before, please provide a link to what you are referring to.

[1206]

http://www.ni.com/swf/flv/labview/us/vista/vr/  hello, here's the link i am pertaining to. Who's alfa then? Sorry, I just badly need it. 😞 Thanks

[smercurio_fc]

see here: http://forums.ni.com/t5/LabVIEW/Voice-Recognition-with-LabVIEW-and-Windows-Vista/td-p/581702

Please do a search. The above link can be found by a simple search which would have taken you about 3 seconds.

[1206]

Thank you so much. I'm sorry. I was just in a hurry. 😞

[1206]

ahmmmmm sir, I just want to ask if this is applicable to microsoft too? I was wondering if this will be applicable too

[smercurio_fc]

I don't understand your question. The example discussed in that videos uses .NET, which is a Microsoft technology. If that is not what you are asking, please rephrase your question.

[1206]

sorry sir, but don't worry I already have used some of the VIs that I find useful in my project

Thank you so much.! 🙂

[sgsdfgaf]

I'm working on Speech Recognition Project.But I'm stuck at the MFCC.
Till now what i have done is...

• Acquired Speech Signal.Detected it's starting and ending.
• Pre emphasized it and divide it into "20ms frames" with 50% overlap.
• Took FFT of each 20 ms frames and thus generated the Frequency spectrum.
• Created 20 Mel Filter bank filter and multiplied "each frame's frequency response" with melfilter banks to get the log energy of each triangular filter.
• Took DCT of log energy' s obtained from triangular bank filters for each frame and thus generated first 12 MFCC for each frame.

My doubt is, now i have 12 MFCC for each frame. Should i add all of them up to get one set of 12 MFCC's?

Was the method done by me for finding the MFCC's correct..?ie taking the MFCC from frequency response of each frame.?

Suppose my method is correct till now.What should i do next.?
I was comparing these MFCC for each frames with MFCC generated from utterance of the same word using euclidean distance and mahalanobis distance.?

Suggestion please.. 🙂

[shahed00012]

Can anyone help to get speech recognition labview function?

I really need it urgent basis.