View Full Version : speech recognition toolkit

April 18th, 2008, 09:24 PM
Hello to everybody!
I am working on a project in my university in which i have to develop a speech recognition program that it will be able to take speech commands. The first part of the project is to name all toolkits (e.g. CMU sphinx) that are able to do that. Do you know anything equivalent to CMU sphinx? Thank you a lot!!!

May 6th, 2008, 11:03 AM
Hi man, I started working on a speech recognition project but I didnīt finished it. I started working with the HMM (Hidden Markov Models), you need to have the models, voice patterns. I can send you some articles and information I have. YOU CAN ALSO GET HTK (Hidden Markov Toolkit), very powerfull tool for speech recognition. I am interested in getting some software for disable people, my email is <snip>. (Mod Note: Email removed to prevent the the email address from getting spammed.)

May 14th, 2008, 10:35 PM
Hi, have you come far with your speech recognition program? I'm kinda interested in this, because I would like to set up a car pc system with built in speech recognition to handle given commands...

May 27th, 2008, 03:12 PM
I planned on cobbling something together with perl/python and Festival/Sphinx for a rather clunky and inelegant example of voice commands and synthesis. Stand by for further details.

June 1st, 2008, 12:10 AM
Litago, unfortunately i haven't done anything yet. I believe that until august i will be able to help you...

June 11th, 2008, 10:07 PM
this was something i was just about to ask everone
nice thread and keep it going

by the way i am also thinking of the same as my colleague project work or some practical marks so can u keep this thing active .....

also small favour please...can u send me all the information u get regarding this project and we shall exchange views abot it what say?

thanks anyway...keep this thread active

June 14th, 2008, 05:19 AM
Hi, have you come far with your speech recognition program? I'm kinda interested in this, because I would like to set up a car pc system with built in speech recognition to handle given commands...

that would be incredibly badass. I am interested in this as well.

July 8th, 2008, 10:38 AM
I have more documentation, I didnīt get far with that project, I was programming a DSP but HHM are a pain in the *** specially when trying to program in C. I recommend you use HTK http://htk.eng.cam.ac.uk/
You donīt have to reinvent the wheel. Now I am using Dragon Naturally Speaking 9, works fine, depends on what kind of project you wanna make.

January 22nd, 2009, 04:25 PM
Hint about HTK: Do not try to learn how to use it
by reading their book. Scroogle "htk tutorial" and
you will find several tutorials which are both
shorter and much more understandable.


January 24th, 2009, 12:37 AM
Just for completeness I think that it is worth a mention that I have had the Microsoft Speech recognition system working within Linux. As yet it has not been developed but it does work and I can provide details of the methods used to get the system functioning. I have just written a server and client for the Microsoft text to speech engine and it would be possible to use this again to implement speech recognition from windows.

Currently the text to speech integrates into speech dispatcher and then uses Orca under Ubuntu to run the whole lot.

I have also got an aMSN plugin that reads all new messages as soon as they arrive along with the speech from Orca.

If you are intrested in more details then let me know.


February 8th, 2009, 04:16 AM
Judging by the lack of fully developed linux open source speech recognition software, i'm guessing that windows software would have to be loaded into wine to get speech recognition on Ubuntu?

Has anyone used any speech recognitiion which is easy to install and use and has good accuracy?

Trying to write a book and it would save my hands and my back if i could lounge back while writing :-)