|
|
|
| Multi-scale Collaborative Coupled Metric Learning Method Based on the Fusion of Class and Structure Information |
| ZOU Guofeng1, FU Guixia1, GAO Mingliang1, YIN Liju1, WANG Kejun2 |
1.College of Electrical and Electronic Engineering, Shandong University of Technology, Zibo 255049
2.College of Automation, Harbin Engineering University, Harbin 150001 |
|
|
|
|
Abstract Aiming at the elements matching problem in different scale space sets, multi-scale collaborative coupled metric learning method based on the fusion of class and structure information is proposed. Firstly, the correlation matrix is constructed under the guidance of class information and structure information of sample distribution. The class information is significant for supervision and the structure information is the auxiliary supervision information. The linear and nonlinear optimization objective equations are constructed based on the correlation matrix. By solving the optimization objective equation, the samples are transformed from different scale space datasets into a unified public space for distance measurement. The experimental results of face recognition show that the nonlinear collaborative coupled metric is an effective measurement method and it is simple and convenient with a higher recognition rate.
|
|
Received: 06 December 2016
|
|
|
| Fund:Supported by National Natural Science Foundation of China(No.61601266), Natural Science Foundation of Shandong Province(No.ZR2016FL14,ZR2015FL029,ZR2015FL034) |
About author:: (ZOU Guofeng(Corresponding author), born in 1984, Ph.D., lecturer. His research interests include metric learning, manifold learning, biometrics and intelligent monitoring.)
(FU Guixia, born in 1985, Ph.D., lectu-rer. Her research interests include intelligent monitoring and intelligent robot.)
(GAO Mingliang, born in 1985, Ph.D., lecturer. His research interests include mo-ving object detection and tracking.)
(YIN Liju, born in 1972, Ph.D., associate professor. Her research interests include space object detection and recognition, digital media technology.)
(WANG Kejun, born in 1962, Ph.D., professor. His research interests include pa-ttern recognition theory and application, biometric identification, intelligent monitoring, calculation of bioinformatics.) |
|
|
|
[1] 沈媛媛,严 严,王菡子.有监督的距离度量学习算法研究进展.自动化学报, 2014, 40(12): 2673-2686.
(SHEN Y Y, YAN Y, WANG H Z. Recent Advances on Supervised Distance Metric Learning Algorithms. Acta Automatica Sinica, 2014, 40(12): 2673-2686.)
[2] DAVIS J V, KULIS B, JAIN P, et al. Information-Theoretic Metric Learning // Proc of the 24th International Conference on Machine Learning. New York, USA: ACM, 2007: 209-216.
[3] XIANG S M, NIE F P, ZHANG C S. Learning a Mahalanobis Distance Metric for Data Clustering and Classification. Pattern Recognition, 2008, 41(12): 3600-3612.
[4] KULIS B, GRAUMAN K, JAIN P. Fast Similarity Search for Lear-ned Metrics. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(12): 2143-2157.
[5] 胡正平,路 亮,许成谦.基于L1范数稀疏距离测度学习的单类分类算法.电子学报, 2012, 40(1): 134-140.
(HU Z P, LU L, XU C Q. L1 Norm Sparse Distance Metric Lear-ning for One-class Classifier. Acta Electronica Sinica, 2012, 40(1): 134-140.)
[6] MOUTAFIS P, KAKADIARIS I A. Semi-coupled Basis and Distance Metric Learning for Cross-Domain Matching: Application to Low-Resolution Face Recognition // Proc of the IEEE International Joint Conference on Biometrics. Washington, USA: IEEE, 2014. DOI: 10.1109/BTAS.2014.6996238.
[7] 陈晋音,何辉豪.基于密度和混合距离度量方法的混合属性数据聚类研究.控制理论与应用, 2015, 32(8): 993-1002.
(CHEN J Y, HE H H. Density-Based Clustering Algorithm for Numerical and Categorical Data with Mixed Distance Measure Methods. Control Theory and Applications, 2015, 32(8): 993-1002.)
[8] 谭飞刚,刘伟铭,黄 玲,等.基于加权欧氏距离度量的目标再识别算法.华南理工大学学报(自然科学版), 2015, 43(9): 88-94.
(TAN F G, LIU W M, HUANG L, et al. Object Re-identification Algorithm Based on Weighted Euclidean Distance Metric. Journal of South China University of Technology(Natural Science Edition), 2015, 43(9): 88-94.)
[9] 赵永威,张 蕾,李弼程,等.基于特征分组与特征值最优化的距离度量学习方法.数据采集与处理, 2015, 30(4): 830-838.
(ZHAO Y W, ZHANG L, LI B C, et al. Distance Metric Learning Based on Feature Grouping and Eigenvalue Optimization. Journal of Data Acquisition and Processing, 2015, 30(4): 830-838.)
[10] LI B, CHANG H, SHAN S G, et al. Coupled Metric Learning for Face Recognition with Degraded Images // ZHOU Z H, WASHIO T, eds. Advances in Machine Learning. Berlin, German: Sprin-ger, 2009: 220-233.
[11] LI B, CHANG H, SHAN S G, et al. Aligning Coupled Manifolds for Face Hallucination. IEEE Signal Processing Letters, 2009, 16(11): 957-960.
[12] 王科俊,阎 涛.核耦合度量学习方法及其在步态识别中的应用.模式识别与人工智能, 2013, 26(2): 169-175.
(WANG K J, YAN T. Kernel Coupled Metric Learning and Its Application to Gait Recognition. Pattern Recognition and Artificial Intelligence, 2013, 26(2): 169-175.)
[13] BEN X Y, MENG W X, YAN R, et al. Kernel Coupled Distance Metric Learning for Gait Recognition and Face Recognition. Neurocomputing, 2013, 120: 577-589.
[14] LI B, CHANG H, SHAN S G, et al. Low-Resolution Face Recognition via Coupled Locality Preserving Mappings. IEEE Signal Processing Letters, 2010, 17(1): 20-23.
[15] ZOU G F, ZHANG Y Y, WANG K J, et al. An Improved Metric Learning Approach for Degraded Face Recognition. Mathematical Problems in Engineering, 2014. DOI: org/10.1155/2014/724978.
[16] ZHANG P, BEN X Y, JIANG W, et al. Coupled Marginal Discriminant Mappings for Low-Resolution Face Recognition. Optik-International Journal for Light and Electron Optics, 2015, 126(23): 4352-4357.
[17] 徐 磊,江 洁,魏振忠.总变分规整化的多帧图像超分辨率重建.电子测量技术, 2012, 35(1): 76-79.
(XU L, JIANG J, WEI Z Z. Multiframe Super-Resolution Reconstruction Algorithm Based on Total Variation Regularization. Electronic Measurement Technology, 2012, 35(1): 76-79.)
[18] 杨艳美,高满囤,贺 剑.维纳滤波图像复原技术的研究与改进.科学技术与工程, 2012, 12(29): 7611-7615.
(YANG Y M, GAO M T, HE J. Wiener Filtering Image Restoration Technology Research and Improvement. Science Technology and Engineering, 2012, 12(29): 7611-7615.) |
|
|
|
2017, Vol. 30 
