关键词: 可视语音, 情感表现力, 基于模型的方法, 基于图像的方法

Abstract:

We summarize and analyze some of the key findings and research methods of the emotional visual speech in recent years.According to different synthesis mechanisms,we first systematically describe and classify the emotional visual speech synthesis technology from the aspects of both image-based methods and model-based methods.And then we compare and discuss their respective advantages,disadvantages,and performance. Moreover,we conduct an indepth discussion on the realization principles of authenticity and the degree of expressive emotion in visual speech synthesis in various literatures. Finally,we point out some other serious problems which should be taken into account,and the direction for further research on emotional visual speech.

Key words: visual speech;expressive emotion;model-based method;image-based method