Designing A Speaker Identification System

   This is my selected BS final project. I wanted to design a system, which identifies its registered users from their voice passwords, which they stated previously. My system primarily should be capable of comparing two pieces of speech and it should have a simple database system for storing and retrieving patterns resulted from theses pieces.
   I first started to search for available scientific resources and free and open source libraries and programs related to this topic. Finally I selected Jialong He's free library for speech/speaker recognition as a base for my work and my final work is a simple text-dependent speaker identification system which uses MFCC features extracted from speech signal as speech patterns and compares them using dynamic time wrapping method. I had placed a link for downloading my project from here before, but it couldn’t be accessed because of netfirms free web hosting limitations (it’s size was larger than 256KB). So I removed that link and intend to prepare it to be downloaded from this page by changing its format in a near future. However if you think you need this program, you can contact me by e-mail.
   You can obtain my programs source code from this page. You can also access a binary version of my programs from this link (Size=241KB) and my project’s Farsi documents from here .