ictsan

Voice recognition in schools was the subject of a study published by the CALL Centre, Edinburgh University in 2003 ; http://callcentre.education.ed.ac.uk/Research/Speech_Recog_PRA/speech_recog_pra.html

The publications included two guides to getting started with voice recognition - each one speacific to the two main software packages at the time; Dragon Naturally Speaking and Viavoice.

At some point since then, Viavoice has been discontinued in terms of new development. It is effectively being marketed as a cheap option. Dragon continues to be developed and is the leader in the field.

There are still a lot of questions hanging over its use by young people and in schools. It has been developed as an alternative for bisiness users to using a secretary. The standard voice files seem to be based on adult males with recived pronunciation (RP) accents. Thus it takes a fair bit of use for the system to become attuned to a youngster with a regional accent.

It has been said that it takes as much time to support a youngster using VR as it does to provide support through scribing for major pieces of writing. It is important to realise;

• The system has to be set up though a period of training in which the user reads aloud from on screen texts. This can last fifteen minutes for a good reader, much longer for a poor reader. An assistant might need to dictate the texts for the poor reader to repeat them.

• The system adjusts to the users voice during each session using the software. Thus if errors in recognition are allowed to go uncorrected, the quality of the recognition deteriorates.

• VR systems are sensitive to background noise

In short, VR will keep on improving as a productivity tool for business users. When it comes to its use by youngsters at home or at shool, it is definitely a high maintenance form of intervention. Think very clearly why you are considering VR. A youngster with motor difficulties might be more effective using word/phrase predicition. A dyslexic with severe spelling difficulties may allow errors and end up with system giving poor recognition. He might be better using speech feedback and quality spellchecking.

A very recent development has been the release of Speak Q; http://www.wordq.com/speakqenglish.html

This is voice recognition aimed at young people. The whole training process is easier and shorter. The youngsters hear the voice and can just reapeat what they hear. There is however, some doubt about how good the voice recognition develop asthe youngster uses it. This is very much a new product. It is more expensive than Dragon. The good news is that a set of demo discs is available. Yes an opportunity to try before you buy.

Recently, a video has appeared on Youtube of a teenage dyslrexic who does use Dragon Voice recognition well;

http://www.youtube.com/watch?v=lXHawlHLmtI