改进非负矩阵分解的神经网络人脸识别

郑明秋; 杨帆

doi:10.3788/YJYXS20173203.0213

您当前的位置：

首页 >

文章列表页 >

改进非负矩阵分解的神经网络人脸识别

显示与图像 | 更新时间：2020-07-07

- 改进非负矩阵分解的神经网络人脸识别
- Face recognition based on improved NMF and neural network
- 液晶与显示 2017年32卷第3期页码：213-218
- 作者机构：
  
  1.长春信息技术职业学院计算机系, 吉林长春 130103
  2.长春理工大学计算机科学技术学院, 吉林长春 130022
- 作者简介：
  
  [ "郑明秋(1977-),女,吉林榆树人,工学学士,讲师,研究方向数字图像处理技术、可视化编程, E-mail:865887390@qq.com" ]
  杨帆(1990-), 男, 安徽阜阳人, 博士研究生,主要从事计算机视觉,图像处理,E-mail：cclg_yangfan@163.com
- 基金信息：
  
  吉林省科技发展计划项目(No.20130303011GX，No.20140204050GX);Supported by Project Agreement for Science & Technology Development of Jilin Province(No. 20130303011GX，No. 20140204050GX)
- DOI：10.3788/YJYXS20173203.0213
  中图分类号： TP391
- 收稿日期：2016-10-08，
  
  录用日期：2016-11-22，
  
  纸质出版日期：2017-03
- 稿件说明：
移动端阅览
郑明秋, 杨帆. 改进非负矩阵分解的神经网络人脸识别[J]. 液晶与显示, 2017,32(3):213-218.

Ming-qiu ZHENG, Fan YANG. Face recognition based on improved NMF and neural network[J]. Chinese journal of liquid crystals and displays, 2017, 32(3): 213-218.
郑明秋, 杨帆. 改进非负矩阵分解的神经网络人脸识别[J]. 液晶与显示, 2017,32(3):213-218. DOI： 10.3788/YJYXS20173203.0213.

Ming-qiu ZHENG, Fan YANG. Face recognition based on improved NMF and neural network[J]. Chinese journal of liquid crystals and displays, 2017, 32(3): 213-218. DOI： 10.3788/YJYXS20173203.0213.

摘要

为了提高人脸识别正确率，提出基于改进非负矩阵分解的神经网络人脸识别算法。首先利用改进的非负矩阵分解对人脸图像进行特征提取，提高非负矩阵分解速度。接着将提取出的特征信息作为神经网络学习入口进行特征训练，由于神经网络在学习过程中，容易出现局部最小值且收敛速度慢等问题，为此采用改进的遗传算法对神经网络进行优化处理，获得最终的人脸识别结果。实验结果表明：利用改进的非负矩阵分解方法能够降低神经网络的分类训练负荷量和运算量，提高人脸识别识别率。通过和各种方法比较可知，本方法的人脸识别率都较高。本方法人脸特征分解速度快，提高了神经网络训练前期精度和收敛速度，使得人脸识别正确率高。当特征向量个数达到40以上时，人脸识别正确率保持95%以上。

Abstract

In order to promote the accuracy of facial recognition

an improved algorithm of the neural network based on NMF for recognizing faces was proposed. First

features from facial images are extracted by using the im proved method of non-negative matrix factorization to increase the decomposition speed of NMF. Then the extracted feature information is made as a neural network learning entrance for characteristics training. Due to the problems of minimum local value and slow convergence speed in the process of neural network learning

an improved genetic algorithm was used to optimize the neural network

and final facial recognition result was obtained. Experimental results indicate that the facial recognition method using the improved NMF can reduce the classification training load and operand of the neural network

and can also increase the facial recognition rate. Compared with other methods

our method has a higher facial recognition rate. The proposed method has high facial feature decomposition speed

it also promote the accuracy of neural network training and improve the convergence speed to make the rate of facial recognition higher. With more than 40 number of feature vectors

the facial recognition accuracy can basically keep more than 95%.

关键词

Keywords

references

FARAJI M R, QI X J. Face recognition under varying illumination based on adaptive homomorphic eight local directional patterns [J]. IET Computer Vision, 2015, 9(3): 390-399.

卢洋,王世刚,赵文婷,等.基于人脸姿态估计的虚拟眼镜试戴技术[J].中国光学,2015,8(4):582-588.

LU Y, WANG S G, ZHAO W T,et al. Technology of virtual eyeglasses try-on system based on face pose estimation [J]. Chinese Optics, 2015, 8(4): 582-588. (in Chinese)

WANG J J Y, BENSMAIL H, GAO X. Multiple graph regularized nonnegative matrix factorization [J].Pattern Recognition, 2013, 46(10): 2840-2847.

TURK M, PENTLAND A. Eigen faces for recognition [J].Journal of Cognitive Neuroscience, 1991, 3(1): 71-86.

HUANG L L, SHIMIZU A, HAGIHARA Y,et al. Face detection from cluttered images using a polynomial neural network [J]. Neurocomputing, 2003, 51: 197-211.

李康顺,李凯,张文生.一种基于改进BP神经网络的PCA人脸识别算法[J].计算机应用与软件,2014,31(1):158-161.

LI K S, LI K, ZHANG W S. PCA face recognition algorithm based on improved BP neural network [J].Computer Applications and Software, 2014, 31(1): 158-161. (in Chinese)

汪海波,陈雁翔,李艳秋.基于主成分分析和Softmax回归模型的人脸识别方法[J].合肥工业大学学报(自然科学版),2015,38(6):759-763.

WANG H B, CHEN Y X, LI Y Q. Face recognition method based on principal component analysis and softmax regression model [J].Journal of Hefei University of Technology, 2015, 38(6): 759-763. (in Chinese)

林庆,李佳,雍建平,等.一种改进的基于NMF的人脸识别方法[J].计算机科学,2012,39(5):243-245,270.

LIN Q, LI J, YONG J P,et al. Improved face recognition method based on NMF [J]. Computer Science, 2012, 39(5): 243-245, 270. (in Chinese)

陈燕芹,段锦,祝勇,等.基于纹理特征的图像复杂度研究[J].中国光学,2015,8(3):407-414.

CHEN Y Q, DUAN J, ZHU Y,et al. Research on the image complexity based on texture features [J]. Chinese Optics, 2015, 8(3): 407-414. (in Chinese)

LI F C, XU L D, JIN C X, et al. Intelligent bionic genetic algorithm (IB-GA) and its convergence [J]. Expert Systems with Applications, 2011, 38(7): 8804-8811.

ZHONG F J, ZHANG J S. Face recognition with enhanced local directional patterns [J]. Neurocomputing, 2013, 119: 375-384.

张岩,武玉强.一种改进的模块PCA人脸识别新方法[J].计算机工程与应用,2011,47(26):216-218.

ZHANG Y, WU Y Q. New face recognition method based on improved modular PCA [J].Computer Engineering and Applications, 2011, 47(26): 216-218. (in Chinese)

李松,刘力军,解永乐.遗传算法优化BP神经网络的短时交通流混沌预测[J].控制与决策,2011,26(10):1581-1585.

LI S, LIU L J, XIE Y L. Chaotic prediction for short-term traffic flow of optimized BP neural network based on geneticalgorithma [J]. Control and Decision, 2011, 26(10): 1581-1585. (in Chinese)

BARTLETT M S, LADES H M, SEJNOWSKI T J. Independent component representations for face recognition [C]//Proceedings of SPIE 3299, Human Vision and Electronic Imaging III. 1998. San Jose, CA: SPIE, 1998: 528-539.

LEE D D, SEUNG H S. Algorithms for non-negative matrix factorization[C]//Advances in Neural Information Processing Systems 14: Proceedings of the 2001 Conference. Vancouver, British Columbia, Canada: MIT Press, 2001, 13: 556-562.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

TFT-LCD表面Mura缺陷的AOI检测研究进展

基于迭代传播网络的快速鲁棒的微光图像增强

基于侧窗滤波改进的神经网络非均匀性校正算法

FRKDNet：基于知识蒸馏的特征提炼语义分割网络

面向光场显示的虚拟视点生成技术进展