Speech Recognition as an Input to Electronic Assistive Technology

Abstract

Automatic speech recognition systems can be enormously beneficial for people with physical disabilities, having the potential to provide a fast and easy-to-use input for computer access and control of the home environment. An overview, based on a review of the research literature and clinical observations, is given of the suitability and performance of speech recognition for computer access in two situations ± for people with no speech problems and for people with dysarthria ± and case studies are presented as examples. Commercial large vocabulary systems are often appropriate for people with no, mild or moderate speech problems, although adequate time, training and support are essential. Those with severe dysarthria achieve lower recognition rates, but speech recognition can still be a useful method of input for some individuals. Speech input as a means of controlling the home environment is mainly based on small vocabulary speech recognition, and recognition reliability is dependent upon the consistency of the speech of the individual. Although speech recognition may be a solution for some people, others reject the technology because of inadequate reliability. This is particularly a problem for people with dysarthria, where speech recognition rates can be unacceptably low due to inconsistent speech.