欢迎访问《图学学报》 分享到:

图学学报

• 计算机仿真与动画 • 上一篇    下一篇

基于加权算法的汉语语音同步三维口型动画研究

  

  • 出版日期:2012-04-27 发布日期:2015-07-28

Research on the Chinese phonetic 3D mouth animation based on weighted control algorithm

  • Online:2012-04-27 Published:2015-07-28

摘要: 目前,国内外对人脸口型动画的研究主要是基于英文开展,而中文人脸口
型动画合成方面的研究还很少。论文结合已有的研究,综合汉语拼音发音时口型变化的规律
和汉语标号的时间控制,提出了声韵加权控制算法,并对整句、整段话中的标号加以权重向
量的分析,能够合成与语音同步的连续变化的三维口型动画模型。在两个连续口型动画过渡
处理上,通过一种余弦函数的插值方法,合成得到两个连续口型之间的过渡口型序列,从而
使合成的汉语语音同步的人脸口型动画更加平滑流畅。

关键词: 汉语语音同步, 三维口型动画, 加权控制算法

Abstract: Presently, the researches on the action of human’s mouth shape in the world are
mainly carried out based on English words. However, such research based on Chinese words is
not so much. Connected with our predecessors’ researches, this paper puts forwards a sound
weighting control algorithm with the integration of the changing law of mouth shape and the time
control of Chinese label when doing Chinese pronunciation. We use this algorithm to analyze the
label in a sentence and a paragraph and compose a 3-dimensional model. This model can
synchronize the continuously changing mouth shape with the sounds. A transition sequence of
mouth shapes between two continuous mouth shapes is composed with the help of an interpolation
method using cosine function. Thus, the action of human mouth synchronized with Chinese
phonetics becomes more smooth and fluent.

Key words: Chinese phonetic synchronization, 3D mouth animation, weighted control algorithm