METHOD AND SYSTEM OF SYNTHESIZING EMOTIONAL SPEECH BASED ON PERSONAL PROSODY MODEL AND RECORDING MEDIUM

摘要

PURPOSE: A method for synthesizing an emotional speech based on a personal prosody model, an apparatus thereof, and recording medium are provided to synthesize utterance based on a modeling result by modeling an individual emotional prosody structure. CONSTITUTION: An emotional prosody structure feature extractor(810) analyzes a personal voice and extracts the property of an individual emotional prosody structure. Emotional voice database(890) stores the individual emotional prosody structure. If input text and target emotion are received, an emotion voice mixer(850) converts the input text into non-emotional voice and searches for the personal emotion prosody structure in the emotion voice database. [Reference numerals] (810) Emotional prosody structure feature extractor; (830) TTS system; (850) Emotional voice mixer; (860) Syllable level correcting unit; (870) Intonational phrase level correcting unit; (880) Sentence level correcting unit; (890) Emotional voice database; (AA) Data set; (BB) Personal emotional prosody structure; (CC) Input text; (DD) Target emotion; (EE) Neutral voice; (FF) Emotional voice