计算机应用 ›› 2017, Vol. 37 ›› Issue (5): 1413-1418.DOI: 10.11772/j.issn.1001-9081.2017.05.1413

• 人工智能 • 上一篇    下一篇

间距判别投影及其在表情识别中的应用

甘炎灵, 金聪   

  1. 华中师范大学 计算机学院, 武汉 430079
  • 收稿日期:2016-11-01 修回日期:2016-12-19 出版日期:2017-05-10 发布日期:2017-05-16
  • 通讯作者: 金聪
  • 作者简介:甘炎灵(1988-),女,广西岑溪人,硕士研究生,主要研究方向:数据降维、模式识别;金聪(1960-),女,湖北武汉人,教授,博士,主要研究方向:图像处理、智能信息处理。
  • 基金资助:
    国家社会科学基金资助项目(13BTQ050)。

Margin discriminant projection and its application in expression recognition

GAN Yanling, JIN Cong   

  1. School of Computer, Central China Normal University, Wuhan Hubei 430079, China
  • Received:2016-11-01 Revised:2016-12-19 Online:2017-05-10 Published:2017-05-16
  • Supported by:
    This work was supported by the National Social Science Foundation of China (13BTQ050).

摘要: 针对全局降维方法判别信息不足,局部降维方法对邻域关系的判定存在缺陷的问题,提出一种新的基于间距的降维方法——间距判别投影(MDP)。首先,根据类的中心均值的异类近邻关系定义描述类边缘的边界向量;在这个基础上,MDP重新定义类间离散度矩阵,同时,使用全局的方法构造类内离散度矩阵;然后,MDP借鉴判别分析思想建立衡量类间距的准则,并通过类间距最大化增强样本在投影空间中的可分性。对MDP在人脸表情数据库JAFFE和Extended Cohn-Kanade上进行表情识别实验,并且跟传统的降维方法主成分分析(PCA)、最大间距准则(MMC)和边界Fisher分析(MFA)进行对比,实验结果表明,所提算法能够有效提取更具区分性的低维特征,比其他几种方法分类精度更高。

关键词: 降维, 间距, 判别投影, 类间离散度, 类内离散度, 表情识别

Abstract: Considering that global dimensionality reduction methods lack useful discriminant information, and local dimensionality reduction methods have defects in measuring neighborhood relationships, a novel dimensionality reduction method based on margin, named Margin Discriminant Projection (MDP), was proposed. Depending on the neighbor structure of mean vector of classes, the boundary vector of the class edge was defined by the heterogeneous neighbor relation of the center mean of the class. On this basis, the between-class scatter matrix was redefined, and the within-class scatter matrix was constructed by the global method. The class margin criterion was established based on discriminant analysis, and discriminant information of samples in projection space was enhanceed by maximizing class margin. The expression recognition on JAFFE and Extended Cohn-Kanade data sets presented the comparison of MDP with PCA (Principal Component Analysis), MMC (Maximum Margin Criterion) and MFA (Marginal Fisher Analysis), and the experiment results show that the proposed method can extract more distinguishable low-dimensional features with relatively higher efficiency, and MDP has better classification accuracy than the other methods.

Key words: dimensionality reduction, margin, discriminant projection, between-class scatter, within-class scatter, expression recognition

中图分类号: