基于集成分类非定标摄像机单目表情识别的研究  

Research on Monocular Expression Recognition Based on Integrated Classification Uncalibrated Camera

作　　者：余涛 邹建华[2,3] 徐君 龙卓群 Yu Tao;Zou Jianhua;Xu Jun;Long Zhuoqun(School of Electronic Engineering,Xi’an Aeronautical Institute,Xi’an 710077,China;State Key Laboratory for Systems Engineering,Systems Engineering Institute,Xi’an Jiaotong University,Xi’an 710049,China;Guangdong Xi’an Jiaotong University Academy,Foshan 528300,China)

机构地区：[1]西安航空学院电子工程学院,西安710077 [2]西安交通大学系统工程研究所系统工程国家重点实验室,西安710049 [3]广东顺德西安交通大学研究院,广东佛山528300

出　　处：《黑龙江科学》2023年第20期32-37,共6页Heilongjiang Science

基　　金：广东省佛山市重点领域科技攻关项目(2020001006827)。

摘　　要：提出一个从一非定标摄像机单目视频中对人的面部表情自动识别的方法,通过一种类似面部肌肉分布的弹性模板来获得表情特征,完成相关规则化处理,随着一个表情的产生,这些特征在时空中变化形成的相应时间序列被逐行排列成矩阵,这个矩阵通过一种保持邻域嵌入(NPE)的流行学习方法被降维,通过一种集成分类器HCRF+SVM来识别这种包含表情特征信息的矩阵,利用相关数据库进行测试。结果表明,该方法比用单个分类器(HCRF或SVM)直接从二维面部表情进行特征识别效率更高,比典型的Kotsia方法更鲁棒。An automatic recognition framework for human facial expressions from a monocular video with an uncalibrated camera is proposed.The expression characteristics are first acquired from a kind of deformable template,which is similar to a facial muscle distribution.After associated regularization,the time sequences from the trait changes in space-time under complete expressional production are then arranged line by line in a matrix.Next,the matrix dimensionality is reduced by a method of manifold learning of neighborhood-preserving embedding.Finally,the refined matrix containing the expression trait information is recognized by a classifier that integrates the hidden conditional random field(HCRF)or support vector machine(SVM).The proposed method shows a comparatively higher recognition rate than the individual HCRF or SVM methods in direct recognition from two-dimensional human face traits,and it is shown to be more robust than the typical Kotsia method.

关 键 词：集成分类器 时间序列 面部表情识别 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]