作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程

• 人工智能及识别技术 • 上一篇    下一篇

一种有监督的线性降维人脸识别算法

郭 丽1,郑忠龙1,2,贾 炯1,张海新1,付芳梅1   

  1. (1. 浙江师范大学数理与信息工程学院,浙江 金华 321004;2. 加州大学,美国 默塞德 95348)
  • 收稿日期:2012-10-17 出版日期:2013-11-15 发布日期:2013-11-13
  • 作者简介:郭 丽(1989-),女,硕士研究生,主研方向:模式识别;郑忠龙(通讯作者),教授、博士;贾 炯,教授;张海新、 付芳梅,硕士研究生
  • 基金资助:
    国家自然科学基金资助项目(61170109, 61100119, 11001247);浙江省科技厅基金资助项目(2012C21021)

A Supervised Linear Dimensionality Reduction Algorithm for Face Recognition

GUO Li 1, ZHENG Zhong-long 1,2, JIA Jiong 1, ZHANG Hai-xin 1, FU Fang-mei 1   

  1. (1. College of Mathematics, Physics and Information Engineering, Zhejiang Normal University, Jinhua 321004, China; 2. University of California, Merced 95348, USA)
  • Received:2012-10-17 Online:2013-11-15 Published:2013-11-13

摘要: 保局投影(LPP)忽略了数据的类别标记信息且鲁棒性较差,为此,提出一种线性判别投影(LDP)算法。引入类间权重矩阵和类内权重矩阵,使各流形间的分离性最大,局部子流形的内在紧致性最小,同时通过一种鲁棒的类内处理方式使算法对outlier数据具有鲁棒性。在ORL、AR和Extended Yale B人脸数据集上进行实验,结果表明,与PCA、LDA、LPP、LSDA和LPDP算法相比,该算法的最佳平均识别率较高,分别可达95.3%、93.64%和96.28%,证明了算法的有效性和可靠性。

关键词: 降维, 流形学习, 判别投影, 有监督学习, 保局投影, 线性判别分析

Abstract: Because Locality Preserving Projection(LPP) ignores the label information of the data and it is lack of robustness, this paper proposes a Linear Discriminant Projection(LDP) algorithm. By introducing between-class weight matrix and within-class weight matrix, LDP maximizes the separability of different submanifolds and minimizes the compactness of local submanifolds. Moreover, LDP is robust to outlier data by a robust within-class processing way. Compared with PCA, LDA, LPP, LSDA, LPDP, the experimental results on ORL, AR and Extended Yale B face databases show that the best average recognition rates of LDP are higher, which can reach 95.3%, 93.64% and 96.28%, and this verifies the efficiency of the proposed algorithm.

Key words: dimensionality reduction, manifold learning, discriminant projection, supervised learning, Locality Preserving Projection (LPP), Linear Discriminant Analysis(LDA)

中图分类号: