摘要
与其他言语合成技术相比较，规则合成有两个明显的优点：语音库占用的内存很小和可以灵活控制合成言语的声学特征和韵律特征。本文介绍第一作者在１９８６—１９９０年期间在英国伦敦大学所开发的汉语按规则合成系统，和１９９２年在中国科学院合作期间所作的改进。包括汉语按规则合成的原理和策略，特别是韵律控制的方法，以及对系统的评价。
Comparing with other speech synthesis techniques, the synthesis-by-rule method has two distinguished features'it requires much less storage space for the phonetic library and offers greater flexibility incontrolling acoustic and prosodic parameters. This paper introduces a Chinese speeCh synthesis-by-rule system developed by the first author during 1986 to 1990 at the University College London and improved by authors in 1992 at Academia Sinica, presents the principle and strategies of synthesizing Chinese speech by rulewith speCial emphasis on prosody control, and reports the evaluation of the system.
出处
《声学学报》
EI
CSCD
北大核心
1995年第2期146-155,共10页
Acta Acustica
