人脸图像的身份和表情同步识别方法研究

项目名称： 人脸图像的身份和表情同步识别方法研究

项目编号： No.61201444

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 电子学与信息系统

项目作者： 周晓彦

作者单位： 南京信息工程大学

项目金额： 24万元

中文摘要： 面孔身份和面部表情的同步识别研究是面孔识别研究中富有挑战性的研究课题。由于在许多应用场景中需要同时知道面孔的身份信息和表情信息，因而解决面孔身份和面部表情的同步识别问题不仅具有重要的理论意义，同时具有重要的现实意义。另外，面孔身份和面部表情的同步识别还能充分挖掘二者之间的内在联系，提高识别率。本项目旨在申请人原有面孔识别研究工作的基础上，深入开展基于双线性模型的面孔身份和面部表情同步识别的理论和方法研究。本项目的研究内容包括三大部分：（1）模拟人脑对面孔识别的认知模型，提出基于双线性模型的面孔身份和面部表情同步识别方案；（2）基于该识别方案，提出面孔身份和面部表情同步识别方法；（3）构建一套完整的面孔身份和面部表情同步识别原型系统，该系统包括人脸图像的输入、人脸检测、特征提取以及面孔身份和面部表情同步识别的过程。

中文关键词： 人脸身份识别；面部表情识别；双线性模型；多标签判别分析；迁移学习

英文摘要： The simultaneous recognition of face identity and facial expression is a very challenging research topic. In many scenarios, we may need to know both face identity and facial expression, the research of simultaneous face and facial expression recognition has been very valuable, not only in theoretical aspect but also in practical applications. On the other hand, it would be helpful to improve the recognition rates of both face and facial expression by making use of the internal relationship between them. To this end, in this project we will focus our attention on the research of simultaneous face and facial expression recognition based on our preliminary work in this area. The research can be divided into three parts: (1) In the first part, we will propose to use the bilinear model for describing the simultaneous recognition framework based on human's cognition model; (2) The second part is devoting to the presentation of the simultaneous recognition approaches based on the results in the first part; (3) The last part is to construct a set of complete prototype system which includes the input of face image, face detecting, feature extraction and simultaneous recognition of face and facial expression.

英文关键词： face identity recognition；facial expression recognition；bilinear model；multi-label discriminant analysis；transfer learning