|Table of Contents|

[1] Zhao Yingnan, , He Xiangjian, et al. Enhanced kernel minimum squared error algorithmand its application in face recognition [J]. Journal of Southeast University (English Edition), 2016, 32 (1): 35-38. [doi:10.3969/j.issn.1003-7985.2016.01.007]
Copy

Enhanced kernel minimum squared error algorithmand its application in face recognition()
增强KMSE及人脸识别应用
Share：

Journal of Southeast University (English Edition)[ISSN:1003-7985/CN:32-1325/N]

Volumn:
32
Issue:
2016 1
Page:
35-38
Research Field:
Computer Science and Engineering
Publishing date:
2016-03-20

Info

Title:
Enhanced kernel minimum squared error algorithmand its application in face recognition
增强KMSE及人脸识别应用
Author(s):
Zhao Yingnan1;  2;  3;  He Xiangjian3;  Chen Beijing1;  2;  Zhao Xiaoping1;  2
1Jiangsu Engineering Center of Network Monitoring, Nanjing University of Information Science and Technology, Nanjing 210044, China
2School of Computer and Software, Nanjing University of Information Science and Technology, Nanjing 210044, China
3School of Computing and Communications, University of Technology, Sydney, Sydney 2007, Australia
赵英男1;  2;  3;  何祥健3;  陈北京1;  2;  赵晓平1;  2
1南京信息工程大学江苏网络监控工程中心, 南京 210044; 2南京信息工程大学计算机与软件学院, 南京 210044; 3悉尼科技大学计算机与通信学院, 悉尼 2007
Keywords:
minimum squared error;  kernel minimum squared error;  pattern recognition;  face recognition
最小均方误差;  核最小均方误差;  模式识别;  人脸识别
PACS:
TP391
DOI:
10.3969/j.issn.1003-7985.2016.01.007
Abstract:
To improve the classification performance of the kernel minimum squared error(KMSE), an enhanced KMSE algorithm(EKMSE)is proposed. It redefines the regular objective function by introducing a novel class label definition, and the relative class label matrix can be adaptively adjusted to the kernel matrix. Compared with the common methods, the new objective function can enlarge the distance between different classes, which therefore yields better recognition rates. In addition, an iteration parameter searching technique is adopted to improve the computational efficiency. The extensive experiments on FERET and GT face databases illustrate the feasibility and efficiency of the proposed EKMSE. It outperforms the original MSE, KMSE, some KMSE improvement methods, and even the sparse representation-based techniques in face recognition, such as collaborate representation classification(CRC).
为了提高核最小均方误差(KMSE)方法的识别能力, 提出一种增强KMSE方法(EKMSE).该方法重新定义KMSE目标函数, 引入一个新的类别标签定义, 并使该定义下的类别标签矩阵能够随核矩阵自适应调整.与通常的目标函数相比, 它能够使不同类别之间的距离增大, 进而提高识别率.同时该算法在参数搜索中采用了迭代技术, 有效提高了算法的计算效率.在FERET和GT人脸库上进行了充分的实验, 结果表明EKMSE算法可行有效.该算法不仅优于原MSE, KMSE以及KMSE改进算法, 也优于目前脸识别中的基于稀疏算法的最新技术CRC算法.

References:

[1] Muller K, Mika S, Ratsch G, et al. An introduction to kernel-based learning algorithms [J]. IEEE Transactions on Neural Networks, 2001, 12(2): 181-202. DOI:10.1109/72.914517.
[2] Xu Jianhua, Zhang Xuegong, Li Yanda. Kernel MSE algorithm: a unified framework for KFD, LS-SVM and KRR [C]//IEEE International Joint Conference on Neural Networks. Washington, DC, USA, 2001:1486-1491.
[3] Xu Yong, Zhang David, Yang Jian, et al. A two-phase test sample sparse representation method for use with face recognition [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2011, 21(9): 1255-1262.
[4] Zhang Lei, Yang Meng, Feng Xiangchu. Sparse representation or collaborative representation: Which helps face recognition? [C]//IEEE International Conference on Computer Vision. Barcelona, Spain, 2011: 471-478.
[5] Qi Zhu. Reformative nonlinear feature extraction using kernel MSE [J]. Neurocomputing, 2010, 73(16/17/18): 3334-3337. DOI:10.1016/j.neucom.2010.04.007.
[6] Zhao Yongping, Sun Jianguo, Du Zhonghua, et al. Pruning least objective contribution in KMSE [J]. Neurocomputing, 2011, 74(17): 3009-3018. DOI:10.1016/j.neucom.2011.04.004.
[7] Zhao Yongping, Wang Kangkang, Liu Jie, et. al. Incremental kernel minimum squared error(KMSE)[J]. Information Sciences, 2014, 270: 92-111. DOI:10.1016/j.ins.2014.02.117.
[8] Xu Yong, Yang Jingyu, Jin Zhong, et al. A learning approach to derive sparse kernel minimum square error model [C]//IEEE International Conference on Control and Automation.Guangzhou, China, 2007: 1278-1283.
[9]Wang Jinhua. A novel solution scheme for the kernel MSE model[C]//International Conference on Artificial Intelligence and Computational Intelligence. Shanghai, China, 2009: 375-378.
[10] Counterdrug Technology Development Program. The FERET database[EB/OL].(2004-06-16)[2016-01-30]. http://www.itl.nist.gov/iad/humanid/feret.
[11] Geogia Institute of Technology. Georgia Tech face database[EB/OL].(2010-01-01)[2016-01-30]. http://www.anefian.com/research/face_reco.htm.

Memo

Memo:
Biography: Zhao Yingnan(1973—), female, doctor, associate professor, ann_zhao_99@163.com.
Foundation items: The Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD), the National Natural Science Foundation of China(No.61572258, 61103141, 51405241), the Natural Science Foundation of Jiangsu Province(No.BK20151530), Overseas Training Programs for Outstanding Young Scholars of Universities in Jiangsu Province.
Citation: Zhao Yingnan, He Xiangjian, Chen Beijing, et al. Enhanced kernel minimum squared error algorithm and its application in face recognition[J].Journal of Southeast University(English Edition), 2016, 32(1):35-38. DOI:10.3969/j.issn.1003-7985.2016.01.007.
Last Update: 2016-03-20