by
Charles Lowe
Florida
State University
Currents in Electronic Literacy Spring 2001 (4), <http://www.cwrl.utexas.edu/currents/spr01/lowe.html>
Endnotes
1. Speech recognition systems are either speaker dependent or speaker independent or a combination of both. In a speaker dependent system the speaker must provide a speech pattern for every word in the vocabulary by training the software, which, considering the size of an English speaker’s vocabulary, would be a long and lengthy process. In speaker independent systems, the software comes with a preexisting vocabulary with speech patterns already provided; yet, differences in speaker dialects and ranges of vocabulary (often determined by the application of the software–doctors and lawyers use a quite different vocabulary subset of the English language than an average user conversing by email) normally require modifications of the vocabulary. Generally, the speech recognition systems today use a common of both, allowing the user to modify a preexisting vocabulary as needed to match the characteristics of the users voice and to add vocabulary for the users particular needs.