当前位置>主页 > 期刊在线 > 信息技术 >

信息技术

基于CNN神经网络的人脸识别模型研究
尚世锋,曹洁,张增
摘要点击次数:147    

摘  要:本文利用卷积神经网络理论,提出一种对样本集进行一种局部特征与全局特征融合训练的方法,达到对于人脸识别的一种方法,通过模型训练,该方法能够最大程度提取人脸关键部位如嘴巴、鼻子、眼睛等部位特征,从而在提取不同人脸特征时,可以更好的从人脸的这些部位中得到更明确的特征向量。


关键词:神经网络;模式识别;人脸识别


作者介绍:

尚世锋(1977-),男,河南许昌人,陆军装甲兵学院讲师,博士。研究方向:战术通信系统。


Research onFace Recognition Model Based on CNN Neural Network

SHANG Shifeng,CAO Jie,ZHANG Zeng

(Department of Informationand Communication,Armored Forces Academy of PLA Army,Beijing 100072,China)

AbstractThe paper based onthe theory of convolutional neural network,puts forward a method oflocal features and global features of the integration of the training sampleset,reach to a method of face recognition,by the training model,this method canmaximize the extraction of face key parts such as nose,eyesand mouth features and other parts,resulting indifferent extraction face the characteristics,can beobtained from these parts of the face feature vector in more clearly.

Keywordsneutralnetwork;pattern recognition;face recognition


参考文献:

[1] http://www.jdl.ac.cn/project/faceId/res-identify.htm.

[2] Y.Bengio,A.Courville,and P.Vincent.Representationlearn ing:a review and new perspectives [J].IEEE transactions on pattern analysis and machine intelligence,2013,35(8):1798–1828.

[3] D.Chen,X.Cao,F.Wen,andJ.Sun.Blessing of dimensional ity:High-dimensionalfeature and its efficient compression for face verification (C).Computer Vision and Pattern Recognition 2013:3025–3032.

[4] J.Masci,U.Meier,D.Cireşan,and J.Schmidhuber.Stacked convolutional auto-encoders forhierarchical feature extraction [J].Artificial Neural Networks and MachineLearning.2011(6791):52-59.

[5] N.Dalal and B.Triggs.Histogramsof Oriented Gradients for Human Detection [C].IEEE Computer Society Conferenceon Computer Vision and Pattern Recognition,2005(1):886–893.

[6] J.Yang,J.Yang,D.Zhang,and J.Lu.Feature fusion:parallel strategy vs.serial strategy [J].Pattern Recognition,2003,36(6):1369-1381.

[7] 杨健,杨静宇,王正群等.一种组合特征抽取的新方法 [J].计算机学报,2002,25(6):570-575.

[8] P.N.Belhumeur,J.P.Hespanha,and D.J.Kriegman.Eigen faces vs.Fisherfaces:recognitionusing class specific linear projection [J].IEEE transactions on patternanalysis and ma chine intelligence,Jul.1997,19(7):711-720.

[9] A.Hyvärinen and E.Oja.Independentcomponent analysis:algorithms and applications [J].NeuralNetworks,2000,13(6):411-430.

[10] J.Zou,Q.Ji,and G.Nagy.A comparative study of local matching approach for facerecognition [J].IEEE Trans.Image Processing,2007,16(10):2617-2628.

[11] Wolf,L.,Hassner,T.,Taigman,Y.Descriptorbased methods in the wild [C].In:Workshop on Faces inReal-Life Images:Detection,Alignment,and Recognition.(2008).

[12] 梁淑芬,刘银华,李立琛.基于LBP和深度学习的非限制条件下人脸识别算法 [J].通信学报,2014,35(6):154-160.