People with disabilities face major problems in their daily lives in communicating with other people. Sign Languages (SLs) are the basic means of communication between hearing-impaired people and also form their natural way of speaking. Systems that could act as interpreters between deaf people and people that do not speak SL, would facilitate the formers? life. Such systems would have to cope with bidirectional translation of sign language sentences and spoken language sentences.

Various research works have been presented recently, concerning mostly SL recognition, rather than spoken language interpretation. The system we present in this paper, aims at Greek Sign Language (GSL) recognition, too, through the use of Hidden Markov Models (HMMs). The recognition rates we achieved for GSL sentences, formed out of a 33-sign vocabulary, exceed 86% and are quite promising.