文本驱动的唇动合成系统  

Text-Driven Lip Motion Synthesis System

作　　者：晏洁[1] 

机构地区：[1]哈尔滨工业大学计算机科学系

出　　处：《计算机工程与设计》1998年第1期31-34,共4页Computer Engineering and Design

摘　　要：国内在汉语拼音到口型的变换方面还没有相应的研究。本文根据汉语拼音的构成及发音时唇动变动特点，首先定义了一个基本口型集，其中包括６种基本口型，再根据该基本口型集，衍生出一个韵母口型库，使每一个汉字在发音时都对应着相应的口型。文本驱动的唇动合成系统的功能为对于任意输入文本，系统均可将该输入文本切分为单独的汉字并对应到三维虚拟人脸的唇动变化中去。本系统的实现在聋哑人辅助教学，提高聋哑人与正常人之间日常交流方面进行了有益的尝试。No internal research has been done in the aspect of transformation from Chinese phoneme to lip motion. In this paper, a basic lip shape set is set up, which consists of six basic lip shapes, and is based on the construction and lip motion characters of Chinese phoneme. Then a base of lip shape rhyme is constructed according to the basic lip shape set, thus every Chinese word is assigned a lip shape. This system accomplishes the transformation of Chinese words from the random input text to the lip motion of 3-D virtual face. This system makes a useful attempt to achieve deaf-mute aid education and enhances daily communication between the deaf-mute and the normal.

关 键 词：语音学 文本驱动 唇动合成系统 CAI 

分 类 号：G434[文化科学—教育学]