High Performance Real-Time Speech Recognition
25 October 1988
Automatic speech recognition provides people with the ability to communicate with machines by voice rather than by keyboard entry or touch panels. As in all fields, there are some problems whose solutions are within our grasp today and others that that require years of additional research. In the near-term, new technology is emerging that can function successfully in applications characterized by small to medium vocabulary spoken in fluent sentences by a varied speaker population. Such technology promises enhanced communications with machines in many applications where a person's hands are otherwise occupied, or even in situations where keyboards are just inconvenient.