Implementation of speaker identification system by means of personal computer

摘要

Speech processing systems are highly complex and teaching students in this subject matter with the underlying technologies can be a challenging task. The aim of this work was to give a hands-on experience via a development of speech processing system based on the hidden Markov model (HMM) as a teaching aid. A method for implementing the speaker recognition system using our toolkit was developed as a dedicated laboratory environment for students. For speaker recognition, experiments were performed to evaluate the performance of the system with 30 speakers (22 impostors and 8 clients). The identification error was 2%, the false acceptance rate was 28% and the false rejection rate was 1%. The Teaching Module Speech Recognition (TMSR) toolkit was used in the lab which was part of the courses on digital signal processing (DSP) technology given by the Computer Engineering and Microelectronics Department. Students are given some initial guidance on how to use the toolkit and instructions to carry out the speaker identification experiments. Overall, the laboratory system was a success and plans are taken in the coming academic years to improve and extend the capability of the system