表情识别研究方法及发展  

Research Methods for Facial Expression Recognition and Its Development

作　　者：文萍[1,2] 侠牧[2] 

机构地区：[1]广西师范大学职业技术师范学院,广西桂林541004 [2]广西师范大学教育科学学院,广西桂林541004

出　　处：《广西师范大学学报（哲学社会科学版）》2012年第4期91-95,共5页Journal of Guangxi Normal University(Philosophy and Social Sciences Edition)

摘　　要：表情是情绪清晰的外显信息。表情识别研究可以了解人类情绪的奥秘,其不断取得的进展得益于表情识别研究方法的发展。表情识别的经典研究方法包括基本判别法、倒脸识别法、局部识别法和表情搜索任务。计算机技术的进步一方面使得表情刺激更容易制作和更具有生态学效度,另一方面产生了微表情识别任务和混合识别法等新方法。表情识别的研究方法正在沿着模拟现实和揭露特殊规律这两种设计思路不断发展。Facial expression is a clear external display of emotions. Secrets of human emotions can be understood through research of facial expression recognition. And the research of facial expression recognition benefits from its development of research methods. The classical methods of facial expression recognition include basic judgment recognition, inverted face recognition, partial recognition and facial expression search task. Benefiting from the development of computer science, icons of facial expression are becoming easier to make, and some new methods, such as the micro-expression recognition task and the mixed recognition task, have been produced. Along with the virtual reality and the exposure of special rules, the research methods for facial expression recognition are developing continually.

关 键 词：表情识别 研究方法 发展 模拟现实 特殊规律 

分 类 号：B844.2[哲学宗教—发展与教育心理学]