A MODULAR APPROACH TO CUSTOMIZED SYNTHETIC VOICES FOR THE NON-SPEAKING

THE NONSPEAKING SUFFER FROM A LACK OF NATURAL-SOUNDING SYNTHETIC VOICES THAT ARE APPROPRIATE IN TERMS OF AGE, SEX, AND DIALECT. THE CREATION OF A WIDE RANGE OF NATURAL-SOUNDING VOICES HAS BEEN HAMPERED BY HIGH DEVELOPMENT COSTS RESULTING FROM A LACK OF ADEQUATE LINGUISTIC MODELS AND FLEXIBLE SOFTWARE TOOLS FOR TEXT-TO-SPEECH SYNTHESIS. ELOQUENT TECHNOLOGY, INC. (ETI) PROPOSES A NOVEL MODULAR APPROACH TO RULE-BASED SPEECH SYNTHESIS COUPLED WITH A NEW SYNTHESIS MODEL THAT WILL LEAD TO COST-EFFECTIVE DEVELOPMENT OF HIGH-QUALITY SYNTHETIC VOICES FOR THE NONSPEAKING. THE KEY TOOL IN THIS APPROACH IS DELTA, A PROGRAMMING LANGUAGE DEVELOPED BY ETI FOR EXPRESSING TEXT-TO-SPEECH ALGORITHMS (RULES). IN THE MODULAR APPROACH, A SINGLE PROGRAM MODULE (THE BASE MODULE) BUILDS THE PART OF AN UTTERANCE REPRESENTATION COMMON TO ALLVOICES IN A GIVEN LANGUAGE; SMALLER, INDEPENDENT VOICE MODULES PRODUCE THE VARIATIONS AMONG THE VOICES. THE SYNTHESIS MODEL USES DELTA'S INNOVATIVE "MULTI-STREAM" DATA STRUCTURE FOR REPRESENTING UTTERANCES. IN THIS PHASE I PROJECT, WE WILL TEST THE FEASIBILITY OF THE MODULAR APPROACH BY IMPLEMENTING A BASE RULE SET FOR ENGLISH, AND VOICE MODULES FOR TWO AMERICAN DIALECTS.