期刊文献+

基于状态估计的张量分解人脸识别方法 被引量:1

Face recognition method based on state estimation and tensorfaces algorithm
下载PDF
导出
摘要 张量脸算法是分析和表达多因素影响的人脸图像结构的一种有效的数学模型,然而张量分解对状态空间的非线性处理仍存在不足之处。对此提出了一种新的多姿态人脸图像识别方法,在原有的张量脸算法基础上结合状态估计的方法。将训练样本图库中不同状态的人脸通过PCA分解得到多种状态(角度、光照、表情)分别对应的特征空间,对于测试样本先投影到每个特征空间,利用最近邻分类器进行状态估计,对利用张量脸算法得到的张量脸进行识别。实验结果表明,该特征提取方法的识别率优于原有的张量脸算法。 Tensorfaces algorithm is an effective mathematical model which can analyze and express the frames of multi-view face images,but there are some problems of multi-linear analysis method with nonlinear changes of face images.So an improved tensorfaces algorithm is proposed for multi-view face recognition which integrates state estimation.The train-ing face images from different states are decomposed to some eigenspaces(views,illuminations and expressions) by PCA.Then the testing face images can be projected into each eigenspace and estimate the states of the unknown images by the closest classifier.It can recognize the faces by tensorface of every image which is obtained by the tensorfaces algorithm.Ex-perimental results show that this method outperforms the original tensorfaces method.
出处 《计算机工程与应用》 CSCD 北大核心 2011年第24期143-145,共3页 Computer Engineering and Applications
基金 国家自然科学基金(No.60572034 No.60973094) 2006年教育部新世纪优秀人才计划项目(No.NCET-06-0487) 江苏省自然科学基金(No.BK2006081) 江南大学创新团队研究计划项目(No.JNIRT0702)~~
关键词 张量脸 状态估计 人脸识别 最近邻分类器 tensorfaces state estimation face recognition closest classifier
  • 相关文献

参考文献8

  • 1Turk M,Pentland A.Eigenfaces for recognition[J].Cognitive Neu- roscience, 1991,3 ( 1 ) : 71-86. 被引量:1
  • 2Belhumeur P N,Hespanha J P,Kriegman D J.Eigenfaces vs fish- effaces:recognition using class specific linear projection[J].IEEE Transactions on PAMI, 1997,219(7) :711-720. 被引量:1
  • 3He X, Yah S.Face recognition using Laplacianfaces[J].IEEE Transactions on PAMI,2005,227(3):328-340. 被引量:1
  • 4Vasilescu M A O,Terzopoulos D.Multilinear analysis of images ensembles: tensorfaces[C]//Proceedings of European Conference on Computer Vision.Berlin: Springer-Verlag, 2002,2350: 447-460. 被引量:1
  • 5姜珊,王羽,田春娜.基于张量脸的多姿态人脸识别算法[J].科学技术与工程,2009,9(15):4381-4385. 被引量:2
  • 6Vasilescu M A O, Terzopoulos D.Multilinear subspace analysis of image ensembles[J].Computer Vision and Pattern Recognition,2003, 2: 93-99. 被引量:1
  • 7Vasilescu M A O, Terzopoulos D.Multilinear image analysis for facial recognition[C]//Proceedings of the International Conference on Patter Recogniton, 2002,2: 511-514. 被引量:1
  • 8Lee Hyung-Soo, Kim Daijin.Tcnsor-based AAM with continuous variation estimation: application to variation-robust face rccogni- tion[J].IEEE Transactions on PAMI,2009,31 : 1102-1116. 被引量:1

二级参考文献7

  • 1Vasilescu M A O,,Terzopoulos D.Multilinear image analysis for fa-cial recognition[].Proc of the International Conference on Pattern Rec-ognition(ICPR).2002 被引量:1
  • 2He X,Niyogi P.Locality preserving projections[].advance in neural information processing systemsNIPS.2003 被引量:1
  • 3Bouman C A.Cluster:an unsupervised algorithmfor modeling Gauss-ian mixtures. http://www.ece.purdue.edu/~bouman . 被引量:1
  • 4Alex M,Vasilescu O,Demetri Terzopoulos.Multilinear Analysis of Image Ensembles:TensorFaces[].Proceedings of theth European Conference on Computer Vision.2002 被引量:1
  • 5LATHAUWER L D,MOOR B D,VANDEWALLEJ.A multilinear singualr value decomposition[].SIAMJournal of Matrix Analysis and Applications.2000 被引量:1
  • 6Kolda,T. G.Orthogonal tensor decompositions[].SIAM Journal on Computing.2001 被引量:1
  • 7Chen S,Zhao H,Kong M,et al.2DLPP:A two-dimensional extension of locality preserving projections[].Neurocomputing.2007 被引量:1

共引文献1

同被引文献10

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部