汉语语音同步的三维口型动画研究  

Chinese speech synchronized 3D lip animation

作　　者：米辉辉[1] 侯进[1,2] 李克豹 甘凌云[1] 

机构地区：[1]西南交通大学信息科学与技术学院,成都610031 [2]南京大学计算机软件新技术国家重点实验室,南京210093

出　　处：《计算机应用研究》2015年第4期1244-1247,共4页Application Research of Computers

基　　金：国家自然科学基金面上项目(61371165);浙江大学CAD&CG国家重点实验室开放课题(A1416);计算机软件新技术国家重点实验室开放课题基金资助项目(KFKT2013B22);四川省动漫研究中心2012年度科研项目(DM201204)

摘　　要：针对汉语的发音习惯以及语音可视化技术中对口型动画自然、连续的要求,提出了一种基于肌肉模型与协同发音模型的与语音保持同步的口型动画的方法。首先,根据汉语发音时的口型视位特征将声、韵母音素归类,并用数据映射的方式合成与之对应的口型关键帧。通过分析输入的文本信息,合成与语音保持同步的三维人脸口型动画。为了解决汉语发音习惯的问题,设计了一种基于微分几何学描述的协同发音建模的方法,该方法通过分析相邻音子间视素的影响权重,可以产生符合汉语发音习惯的口型动画。最后,通过实验对比和分析,该方法产生的口型动画更为逼真,且符合汉语发音的习惯。In order to meet the characteristics of Chinese pronunciation and satisfy the requirement of the speech visualization technology,namely the natural and continuous lip animation,this paper proposed a method of synchronized speech 3D lip animation based on muscle model and coarticulation model. Firstly,according to the mouth visemes characteristics of the Chinese pronounciation,consonant-vowel phonemes were grouped and Chinese visual phoneme key frames were synthesized by data mapping. And then,the method depended on the analysis of the input Chinese text to simulate the synchronized speech 3D facial animation. Moreover,a coarticulation model,which conformed to geometric description,was constructed where the pronunciation property of Chinese characteristics would be sufficiently considered. This method used the inter syllables weighting function of consonant-vowel to simulate the effect of coarticulation and lip animation that obeyed Chinese pronunciation habit.Finally,to testify the performance,it simulated the 3D facial lip animation. The results show that the synthesized lip animation is more natural and accord with the habits of Chinese pronunciation.

关 键 词：语音可视化 协同发音模型 口型动画 语音动画 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]