An Educational Software Tool
for
Speech Coding
in 
MATLAB
Dr. Andreas Spanias and Ted Painter,
ASU Speech and Audio Processing Lab
Last updated: May 22, 1997
 
 
Contents  
Introduction  

What is Speech Coding ?

Speech coding or compression is the field concerned with compact digital representations of speech  signals for the purpose of efficient transmission or storage. The central objective is to represent a signal with a minimum number of bits while maintaining perceptual quality. Current applications for speech and audio coding algorithms include cellular and personal communications networks (PCNs), teleconferencing, desktop multi-media systems, and secure communications.   
Project Description  
 

What educational software tools we are developing in the ASU Speech and Audio Lab? 

We are currently developing educational simulation software which helps students to explore and understand standardized speech coding algorithms.  MATLAB simulations have been created for several internationally standardized algorithms, in particular  The simulation software provides an interactive environment that allows users to display time- and frequency-domain representations of input and reconstructed speech on any MATLAB host computer with a sound card. Scores associated with subjective and objective performance measures are computed using simulation tools. The tools are used in undergraduate and graduate DSP courses to introduce speech coding fundamentals. 
Example Simulations  

What are the capabilities of our tool ? 

The links in this section give sample simulation output and signal viewing windows.  First, the tool presens users with a menu of available algorithms.   The algorithms process input speech frame-by-frame.  For each frame processed by the selected algorithm, users are able to view a time-domain signal representation and a frequency-domain signal representation for both input and output speech. The examples given here are for the G.728 (LD-CELP) coder.  The viewing  windows allow a user to examine and  compare the waveform matching and spectral matching characteristcis of each algorithm.  Other windows are also available.  Most of the speech coders rely upon a source-system signal representation, which is realized as a combination of an excitation sequence (lower vocal tract) and a digital filter (upper vocal tract filter).  The tool allows examination of both the source/excitation, and the system/vocal tract filter.  Other features are also being added, some of which are described under Current Topics. 
Current Topics  

We are presently making several enhancements to the existing system, including:

  • Integration of interactive exercises
  • Integration of bit error (BER) performance testing
  • Integration of transcoding performance testing (tandeming)
  • Integration of objective performance measures
  • Integration of automatic subjective performance measures
  • Inclusion of a low rate Sinusoidal Transform Coder 
  • Inclusion of  recently standardized algortihms 
  • Incorporation of optional links to high-speed compiled coders
  • Improvements to the user interface
People  
 

This software is being developed in the ASU Speech and Audio Processing Lab by: 

Dr. Andreas Spanias 
Ted Painter
Publications  

Some documents are available which give more details on our work: 

  • Spanias, A..S., and Painter, E. M.  (1996). "A MATLAB Software Tool for the Introduction of Speech Coding Fundamentals in a DSP Course."  
  • Painter, E.M., and Spanias, A.S. (1996) "A MATLAB Software Tool for the Introduction of Speech Coding Fundamentals in a DSP Course.",  in FIE '96, Nov. 1996 (Follow the link, then search on-line proceedings for an electronic copy of the paper.)
    		•
    Related Sites  

    Other sites related to this project include the following: 

    Standards Bodies Technical Societies UUCP Newsgroups Companies
    1. ETSI 
    2. ISO 
    3. ITU 
    1. ASA
    2. IEEE Education 
    3. IEEE Sig. Proc. 
    1. comp.speech  
    2. comp.dsp  
    1. ASPI
    2. MathWorks 
    3. TI
    Conferences Demo Software
    1. Frontiers in Education 
    1. LPC-10e
    Table 1  Resources Related to Our Educational Software Tool on Speech Coding
     
    Contacts 

    For further information, direct all correspondance to: 

    Dr. Andreas S. Spanias  <spanias@asu.edu>
                              
    Site Last updated on May 22, 1997