Catalogue of Artificial Intelligence Techniques
Keywords: recognition, speech, subvocal
Author(s): Kevin Dorrian
Subvocal Recognition is the term used to describe the action of interpreting the movement in the articulator muscles in the mouth and jaw and converting this into text. When speaking aloud air is forced up past the larynx and tongue and the words are formed using the articulator muscle. This muscle moves regardless of air being forced past it therefore when reading a word without making sound or mouth movements, using electrodes attached to the throat, weak electrical signals can be detected which can be interpreted as words.
Subvocal recognition is similar to speech recognition in the way it converts speech signals into words. This technique relies heavily on pattern matching and instance based learning to correlate muscle movements with words.
- D. Jurafsky and J.H. Martin, Speech and Language Processing, Prentice Hall, 2000.
- John Bluck, NASA Press Release, March 17, 2004, Nasa, 2004, pp.1.
- David Armstrong, The Silent Speaker, April 10, 2006 Forbes 7 (2006) no.April 10th, Forbes, 1.