History

• Speech recognition was created in 1936 by AT&T's bell labs, but it was not a commercial product. Their invention was called the Voder and functioned using a set of keys and levers. The Voder could emit either pure tone or white noise and would imitate the human voice. The user could also emit inhuman noises if skilled enough.

• The largest speech recognition project was in 1971. It was the Speech Understanding Research by DARPA (Defense Advance Research Projects Agency). A man named Lawrence Roberts spent $15 million on this project, which took 5 years.

• In 1978, Texas Instruments invented the Speak and Spell. It was a talking educational toy that was designed to help children learn the alphabet and spell. It also included Speak and Read and Speak and Math functions.

• Speech recognition wasn't a commercial product until 1982, when Covox added sound to a few pc's. These pc's were the Commodore 64, Atari 400/800, and IBM's pc. With the sound that was added to these pc's, speech recognition was also added to them.

• Currently, the top brand of speech recognition software is made by a company called Dragon Systems, which was founded in 1982. One of their products is Dragon Naturally Speaking.

• For more information on the history of speech recognition see this timeline.