计算机工程 ›› 2011, Vol. 37 ›› Issue (4): 210-211.doi: 10.3969/j.issn.1000-3428.2011.04.075

• 人工智能及识别技术 • 上一篇    下一篇

优化局部保留投影及其在表情识别中的应用

黄 勇   

  1. (柳州铁道职业技术学院电子工程系,广西 柳州545007)
  • 出版日期:2011-02-20 发布日期:2011-02-17
  • 作者简介:黄 勇(1980-),男,讲师、硕士,主研方向:模式识别,自动检测技术

Optimized Locality Preserving Projections and Its Application in Expression Recognition

HUANG Yong   

  1. (Department of Electronic Engineering, Liuzhou Railway Vocational Technical College, Liuzhou 545007, China)
  • Online:2011-02-20 Published:2011-02-17

摘要: 提出一种基于优化局部保留投影(OLPP)的人脸表情识别方法。OLPP方法在降维过程中将图像结构信息融入LPP目标函数,通过降维处理,在获得图像结构信息的同时将投影最优化,从而能从原始表情数据中提取更多更具判决性的有效表情信息。JAFFE和CED- WYU(1.0)2个表情数据库的识别结果表明,基于OLPP的特征提取方法能有效提高识别率。

关键词: 局部保留投影, 优化局部保留投影, 表情识别

Abstract: A facial expression recognition method based on Optimized Locality Preserving Projections(OLPP) is proposed. Unlike Locality Preserving Projections(LPP), OLPP incorporates graph construction into the LPP objective function in dimension reduction process, thus obtains a simultaneous learning framework for graph construction and projection optimization. OLPP can extract more useful and discriminating expression features from the original expression data. Experimental result on JAFFE and CED-WYU(1.0) shows that OLPP is an effective method for improving the recognition accuracy.

Key words: Locality Preserving Projections(LPP), Optimized Locality Preserving Projections(OLPP), expression recognition

中图分类号: