SPEECH RECONVICTION

   

SPEECH RECONVICTION:

        Speech Recognition is a technology that allows a user to use his or her speech to control the computer, or convert spoken words to text. The differentiating factors for different Speech Recognition software are the amount of training needed by the software to recognize a user's speech patterns, the ability of the software to recognized continuous speech, and the number of words the software can recognize. Newer software requires very little training, and can recognize continuous speech with a very large vocabulary. They boast of accuracies as high as 98%. The ability to speak with conviction can be developed through practice and acquiring comfort in public speaking. Good speakers have knowledge of their topics, enthusiasm, clear diction, and volume control. They are good at being versatile and know how to work with different audiences. Also, there is much software that require no training, but can recognize a small number of words, like the digits from zero to nine. The former software are use for converting speech to text, while the later are used by phone companies to give access to menus. Examples of such software are 'Dragon naturally speaking' and 'IBM via Voice'. Microsoft Office XP and 2003 also come with this feature. This technology has been under development since the 1990s, but it not yet sufficiently wide spread.