基于加权DFFD算法和渐变动画思想的人脸动画系统  

Facial Animation System Based on Weighted DFFD Algorithm and Thought of Gradient Animation

作　　者：崔明[1] 许志闻[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012

出　　处：《吉林大学学报（理学版）》2012年第2期288-292,共5页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:60973091);吉林省科技发展计划重点项目(批准号:20080325)

摘　　要：使用加权Dirichlet自由变形算法并结合渐变动画思想的方法实现文字/语音驱动的人脸动画系统,解决了传统方法中建模过程复杂、输出动画不流畅、唇形和语音对应效果差等问题.实验结果表明,应用该方法可得到较真实的实时动画效果.The authors adopted the combination of weightd Dirichle t free-form deformation algorithm and thought of gradient animation to build th e text-driven and speech-driven facial animation system.It has solved many pr oblems in traditional methods such as the complexity of modeling process,the lo w fluency in animation output,the inconsistency between lip and voice and so on.Because of these advantages it achieves a more realistic real-time animation and shows very positive practical and theoretical value as reference.

关 键 词：语音识别 人脸动画 语音驱动 文字驱动 Dirichlet自由变形算法 渐变动画 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]