Richard G Kheir

Advisor: Dr. Thomas Way

Title: Improving Speech Recognition to assist real-time classroom note taking

Description:

Note takers are the only assistance available to help deaf and hard of hearing students to get some information out of a classroom lecture. A note taker however, can not keep up with the lecturer's speech speed and has to chose certain information to write down. This information might not be what the student requiring assistance need to know. Furthermore, the note taker's major should, most of the time, be the same as the student's.

To solve this limitation, we are developing an automatic speech recognition transcriber that uses the Microsoft Speech Recognition Engine along with a distributed java application to transcribe speech. Our aim is to maximize accuracy and transcribtion speed.
 

*******************************************************************

Project Status:

Date Activity Links
12/15/2005 DiBS Application DiBS download page
12/30/2005 RESNA paper draft  
01/05/2006 Paper submitted to RESNA Read the paper here
02/12/2006 Program recognizing speech without a GUI  
03/03/2006 GUI created  
03/23/2006 Tweaked the application to improve accuracy  
05/18/2006 IS extended over the summer  
06/23-06/27 29th Annual RESNA conference - Atlanta, GA Download presentation here (TODO: upload required swf files)
07/09/2006 Finish networking and send application to Dr. Way for testing  
07/30/2006 Submit IS and start with journal/conference paper  
07/18/2006 Program done - Need access to webster to run the server  
     
     
     
Future work Explore the Sphinx engine for portable speech recognition  
     
     
     
     
     

*******************************************************************

Links:

 

*******************************************************************