Speech2sign is an open web platform for developing, compiling, and running rule-based speech to sign language translation applications. Speech recognition is performed using the Nuance Recognizer 10.2 toolkit. Signed output, including both manual and non-manual components, is rendered using the University of East Anglia's JASigning system. Output in the form of recorded video is also supported, with tools that permit efficient management of the video recording process.
Speech2sign is designed to make the component technologies readily accessible to sign language experts who are not necessarily computer scientists. Translation grammars are written in a version of Synchronous Context-Free Grammar adapted to the peculiarities of sign language. All processing is carried out on a remote server, with content uploaded and accessed through a web interface. Our initial experiences show that simple translation grammars can be implemented on a time-scale of a few hours to a few days and produce signed output readily comprehensible to Deaf informants.
The platform has been created in a collaboration between the University of Geneva and the University of Zurich, under funding from the Boninchi, Max Bircher and Raiffeisen Foundations, Credit Suisse and the Loterie Romande. In collaboration with the Teamco foundation.