Advances in Voice Recognition
From: Futurist - 02/2005-  Vol. 39, No. 1, P. 7
By: Janine Lodato
 
Voice recognition technology is imperfect, but an essential tool for people
with disabilities that prevent them from using a keyboard and mouse, writes
Hi-Tech Inventions senior partner Janine Lodato, who has multiple sclerosis
and uses IBM's ViaVoice software. As the technology improves and is augmented
by computer vision capabilities, non-disabled users are likely to abandon
typing text as well. The present state of voice recognition technology
requires patient training of the software and skills development on the part
of the user. In addition, the headset microphones that many disabled users
need for computer dictation interfere with their use of gyroscopes that use
head movements to control mouse functions. Voice recognition relies on
dictation mode, in which spoken words are translated into text, with accuracy
determined by clear enunciation, the audio environment, and the software's
learned knowledge of the user's vocabulary and pronunciation. The command
mode is prompted by a code word, and provides access to basic editing actions
such as capitalization or discrete spelling. Documents created solely with
voice recognition software still often require editing, but the technology
vastly increases productivity for disabled users. Advances in desktop
computing power and peripherals will help, but computer vision technology
that allows lip-reading and eye-tracking will make voice recognition
solutions much more effective in the future.

Read the entire article at:
http://www.wfs.org/trendjf05.htm

Links:
A New Way of Thinking - High-Tech Style
http://www.eclicktick.com/voicerecognition.htm
http://www.disabledperson.com/articles/thinking.asp

Linux-Based Voice Recognition
http://www.linuxjournal.com/article/6610