REAL-TIME SPEECH PLAY SYSTEM AND METHOD THEROF

摘要

PURPOSE: A real-time speech play system and the method thereof are provided to execute speech play and speech synthesis processing appropriately in real-time for a voice information service dependent on a result of speech recognition in a speech processing board into which speech recognition, text to speech and speech play are integrated. CONSTITUTION: A speech processing part(32) composed of a speech recognition module(32a) and a text-to-speech module(32b) executes two services in a board selectively according to a message. A recognition memory part(34) is configured for recognition parameters necessary for recognition and other memory domains to carry out a recognition program. A text to speech memory part(35) is configured for a data base for text to speech and other memory domains to carry out a text to speech program. A common memory part(36) consists of a first synthesis data output domain(36a) and a second synthesis data output domain(36b) for real-time text to speech and speech play. A speech play part(37) regenerates a digital voice data. An external interface connected with communication channels carries out the transmitting and receiving of speech.