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 Tools

Status
Not open for further replies.

ali

Member level 2
Member level 2
Joined
Feb 16, 2003
Messages
43
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Activity points
454
hi,
Do you any free code for speech recognitions ?
 

Dear ali
Try the following Link

**broken link removed**

Also Read the paper attached herewith

I hope it will help you.

Regards
 

hi ali
i wanna build the software for speech recognition
we could start together
i would use matlab
 

Dear freewilly30,
It's a nice idea to make a joint effort. One of my student is also doing his project on Speaker recognition and he is also targetting the matlab. If you like then i can introduce him to you. I think he has done nice work. If you are interested then give me your personal email or some other contact as you like.
And same offer is also for ali...

-Regards
 

Dear freewilly30,
I have added u in my contact list of my office. My name is shafique and now onwards if i will send you any email with the reference of EDA board then do understand.
I need some time to contact my student. As far as i remember our last discussion he was intending to implement the vector quantization and some classifiers in the matlab.

I have downloaded a code from the site
**broken link removed**

If you want to try this at the mean time then have a look on it.
I am on leave from my office in next days of week so hopefully after monday i will send you some information about the work done or the road map for next work
-Regards
 

Hidden Markov Model Toolkit
htk.eng.cam.ac.uk
 

Dear muoinhohn,
Can you please explain a bit more that how will this tool kit help us.
-Regards
 

topic said:
Do you any free code for speech recognitions ?

swahlah said:
Can you please explain a bit more that how will this tool kit help us.

From: [url said:
https://htk.eng.cam.ac.uk/][/url]
The Hidden Markov Model Toolkit (HTK) is a portable toolkit for building and manipulating hidden Markov models. HTK is primarily used for speech recognition research although it has been used for numerous other applications including research into speech synthesis, character recognition and DNA sequencing. HTK is in use at hundreds of sites worldwide.
 

Dear me2please,
Just a little more information.
In order to use the HTK what exactly i have to do.

-Thanks for your kind help
 

Hi
HTK is a tool for speech recognition based on the HMM technique.
There are some functions for speec recognition in HTK.
1) Future vector of speech (MFCC, PLP, LPC., etc.)
2) Training: Baum-Welch, Viterbi
3) Recognition: Viterbi

In other hand, HTK has also the function for language model.

The tool is writen in ANSI C on Linux and Window. The sources are provided, but HTK is not open source. You can use it only for the research aim.
You can registre on site to download it with its manual that is clearly.
Regards
 
Dear muoinhohn,
Your comments are really valuable. i thought that HTK is in matlab but it is in C. Do you know some functions of MATLAB which can help in speech recognition.

-Regards
 

hi!
if you want to develop a product and your tests are much, it is better to use htk.
i have many experiences on it and it's very good tool for speech recognition.
 

Dear muoinhohn,
I am grateful for your kind help.
I have downloaded the matlab toolbox for HMM and now i will see it,
-Regards
 

Hi swahlah,

I hv try to use the https://www.cs.ubc.ca/~murphyk/Software/HMM/hmm.html toolbox, but the log-likelihood becomes positive, and i read from https://www.cs.ubc.ca/~murphyk/Software/HMM/hmm_usage.html#priorcov mentioned that
"Buried inside of KPMstats/mixgauss_Mstep you will see that cov_prior is initialized to 0.01*I. This is added to the maximum likelihood estimate after every M step. To change this, you will need to modify the mhmm_em function so it calls mixgauss_Mstep with a different value. "

my questions is
How to modify the mhmm_em.m and modify it based on what?

Hopefully you can help me on that matter if you hv the same problems before.

Thanks.
 

Hi

Were you able to realize the solution to number recognition or any other application using HMM ?

DO you have any data on comparison of techniques like HMM, LPC which is better when used for Speech recognition.

Thanks
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top