人脸表情动画与语音的典型相关性分析.docx

人脸表情动画与语音的典型相关性分析*?范懿文1,2+, 柳学成1,2, 夏时洪11(中国科学院计算技术研究所前瞻研究实验室北京100190) 2(中国科学院研究生院北京 100049)The Canonical Correlation Analysis of Facial Expression Animation and SpeechFANYi-wen1,2+, LIUXue-cheng1,2, XIAShi-hong11(Advanced Computing Research Laboratory, Institute Of Computing Technology, Chinese Academy Of Sciences, Beijing100190, China) 2(Graduate University of Chinese Academy of Sciences, Beijing 100049, China)+ Corresponding author: E-mail: fanyiwen@ict.ac.cnReceived 2004-00-00; Accepted 2004-00-00 Abstract: Speech driven facial animation technique is an approach to synthesize the corresponding facial animation according to the speech input. The lips motion, as a component of the speech driven facial animation, has already been developed to a relatively mature extent. The facial expression animation, as another component, which plays a very important role of enhancing the realism of facial animation, becomes increasingly valued by researchers recently. Some previous works proposedmethods of speech driven facial motion which reached certain achievement, but they didn’t quantitatively analyze the relationship between facial expression motion and speech, that resulted in the poor expansibility of these methods. In order to solve this problem, this article adopts Canonical Correlation Analysis (CCA) to quantitatively analyze the correlations between facial expression motion and speech, and reaches the intuitive and quantitative conclusions of them. The analysis result shows that the two are strongly correlated and reveals the specific internalrelations between the components of facial expression motion and the acoustic features. This article can be used as theoretical reference and judging criterion for speech driven facial animation technique.Key words: Facial Expression Animation; Canonical Correlation Analysis(CCA) 摘 要:语音驱动的人脸动画技术是以语音为输入，合成与之匹配的人脸动画的技术。唇形动画作为语音驱动人脸动画的一个组成部分，发展至今已较为成熟。而表情动画作为另外一个组成部分，在增加人脸动画逼真性方面起着重要的作用，在近来受到了越来越多的重视。一些工作提出了语音驱动人脸表情动画的方法，并取得了一