Speech Recognition Research
Dr. Andreas Spanias, Min-Tau Lin, and Ajith Mekkoth, Arizona State University (ASU) 
  Last updated: May 21, 1997 
 
 
Contents  
   
Introduction  
 

What is Speech recognition ?

The goal of automatic speech recognition is to develop techniques and systems that enable machine to accept speech input. The problem of speech recognition has been actively studied since the 1950's. Practically, speech recognition will solve problems, improve productivity, and change the way we run our lives. Intellectually, speech recognition holds considerable promise as well as challenges in the years to come for scientists and product developers. The performance of speech recognizers has improved dramatically due to recent advances in speech science and computer technology. With continually improving algorithms and faster computers, it appears that man-machine communication by voice will be a reality within our lifetime.
 
Objectives  
 

The goal of this research is to develop a robust and high perfromance alphabet recognizer under noise and telephone channel conditions. 

In particular, the research objectives are to: 
  • Develop and evaluate a speech feature that is highly robust in mismatched conditions.
  • Find a good unit of speech to model the alphabet letters.
  • Incorporate acoustic knowledge in the recognizer.
  • Improve stop and nasal consonant discrimination.
  • Develop a training algorithm to improve robustness in mismatched conditions.
 
Research Topics  
 
 

Several research topics are currently under study, including: 

  • Development of an algorithm for high performace alphabet recognition.
  • Development of an algorithm for speech recognition system for noisy and telephone channel environments.
 
Recent Accomplishments  
 

Recent accomplishments include the following:

  • A speech recognizer with speaker adaptation function.
  • A speech recognizer for name retrieval application.
  • Speech recognition system under noise conditions. 
  • Telephone E-set letters recognizer 
    		•
     
    Research Group  
     
     

    This page describes speech recognition research conducted by several people, including: 

    Dr. Andreas Spanias, advisor 
    Min-Tau Lin, Ph. D. student 
    Ajith Mekkoth, Ph. D. student (currently at Nokia Mobile Phones)
     
    Publications  
     
        Loizou, P., Mekkoth, A. and Spanias,  A. S. (1995). Telephone Alphabet Recognition for Name Retriveal Application.
        Loizou, P.,and Spanias, A. S. (1995). Automatic Recognition of Syllable-final nasals preceded by /E/.
        Lin, M. and Spanias, A.S. (1995).  Robust Speech Recognition based on Minimum Error Classification and Weighted Projection Measure. 
        Loizou, P. and Spanias, A.S. (1996).  High-performance Alphabet Recognition.
        Lin, M. and Spanias, A.S. (1996).  An improved Approach to Robust Speech Recognition by using Minimum Error Classification.
         
     
    Related Sites  
     

    Other sites related to this project include the following: 

    Telecommunications Research Center 
    College of Engineering and Applied Sciences 
    Arizona State University, which is located in the city of Tempe, Arizona  85287-7206 USA
     
    Contacts  
     

    For further information, direct all correspondance to: 

    Dr. Andreas S. Spanias <spanias@asu.edu>
    Last updated on May 21, 1997