Application of speech synthesis and resynthesis in two speech-communication aids

Two projects are currently investigating the usefulness of synthetic speech in a speech aid. Since the group of speech-impaired people is very diverse, the two projects aim at different target groups. In one project fixed, prestored sentences are used to create a small, easy to operate speech aid, be it with a limited vocabulary. The other project gives the user an unlimited vocabulary by using text-to-speech conversion and diphone concatenation. In both projects prototypes are being evaluated by potential users. This paper describes these two prototypes and the evaluation.