Building highly realistic facial modeling and animation: a survey
Building highly realistic facial modeling and animation: a survey
复制标题
DOI：
10.1007/s00371-007-0175-y
复制
发表时间：
2008-01-01
期刊：
影响因子：
3.5
通讯作者：
Dong, Feng
中科院分区：
文献类型：
--
作者：
Ersotelos, Nikolaos;Dong, Feng
本文对人脸建模和动画技术进行了综述。该调查是从两个不同的角度进行：人脸建模，它涉及如何产生三维人脸模型，和面部动画，它涉及如何合成动态面部表情。为了生成一个单独的人脸模型，我们可以执行一个通用模型的个性化或联合收割机人脸模型从现有的人脸集合。对于面部动画，我们进一步将技术分为基于模拟，性能驱动和基于形状混合的方法。这些技术的优势和弱点，在每个类别进行了讨论，以及这些技术的应用，以各种开发。此外，简要的历史回顾技术的发展。最后讨论了其局限性和未来发展趋势。文章最后给出了结论。
This paper provides a comprehensive survey on the techniques for human facial modeling and animation. The survey is carried out from two different perspectives: facial modeling, which concerns how to produce 3D face models, and facial animation, which regards how to synthesize dynamic facial expressions. To generate an individual face model, we can either perform individualization of a generic model or combine face models from an existing face collection. With respect to facial animation, we have further categorized the techniques into simulation-based, performance-driven and shape blend-based approaches. The strength and weakness of these techniques within each category are discussed, alongside with the applications of these techniques to various exploitations. In addition, a brief historical review of the technique evolution is provided. Limitations and future trend are discussed. Conclusions are drawn at the end of the paper.