Continue to Site

Welcome to EDAboard.com

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.

speech recognition TMS320C6713

Status
Not open for further replies.

ngayle

Newbie level 2
Newbie level 2
Joined
Apr 2, 2008
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
Jamaica
Activity points
1,296
I need some help with a basic speech recognition program for the TMS320c kit. I have a copy of Chassaing's book which covers some of this but i don't have the CD. Has anyone written or knows of where i could get the C code for speech recognition, the code should allow the individual speaking to be recognized.
 

is this for a class?
What method are you using?
vector quantization, probablistic, ....

I actually built a speech recognition system for that exact chipset.
I used VQ. Worked well.
 

peleda said:
is this for a class?
What method are you using?
vector quantization, probablistic, ....

I actually built a speech recognition system for that exact chipset.
I used VQ. Worked well.

This is for a class assignment, I will be using VQ, but I don't know where to start.
 

The joys of 434/586.

Assuming you understand how VQ works, you could try to implement if first on matlab.

Are you doing word verification or people recognition?

1 - take in a training sample of what you want to verify. if you want to verify people, each person should have a bank.
if you want to differentiate words, each words should have a bank.
- to do this, sample the speech, take the fft, compute Mel frequency cepstrum. apply VQ or divide into clusters.

does this help?
 

    ngayle

    Points: 2
    Helpful Answer Positive Rating
The joys of 434/586.

Assuming you understand how VQ works, you could try to implement if first on matlab.

Are you doing word verification or people recognition?

1 - take in a training sample of what you want to verify. if you want to verify people, each person should have a bank.
if you want to differentiate words, each words should have a bank.
- to do this, sample the speech, take the fft, compute Mel frequency cepstrum. apply VQ or divide into clusters.

does this help?

I am also having this project, and the same problem,
In CD_Examples of DSK 6713, it's having folder namely speaker_recognition,
and now I am modifying this folder to make it to speech recognition,
but don't know what is the difference to make change.
Can you give me some advice, Pedela?
 

hello
i hav implemented the speaker recgintion on the matlab and it now i am implementin it ointhe DSK6713 board as i came to know that it is already built in myexamples of the CD for DSK 6713 can any one tell me the basic steps hpw to build the this speech recognition program on board please reply quick
 

[/COLOR]
is this for a class?
What method are you using?
vector quantization, probablistic, ....

I actually built a speech recognition system for that exact chipset.
I used VQ. Worked well.

Hi i also doing the same but i am using mfcc and also vq for speaker/voice recognition.
I came to know u have the vq c code. can please mail me that code it would be a lot of help. My email id is msaashikit03@gmail.com.
Thanx in advance. if anyone has mfcc code please send me.
 
Last edited:

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top