GNN-CRC: Discriminative Collaborative Representation-Based Classification via Gabor Wavelet Transformation and Nearest Neighbor

doi:10.1007/s12204-018-1960-7

Abstract

Abstract: Collaborative representation-based classification (CRC) is a distance based method, and it obtains the original contributions from all samples to solve the sparse representation coefficient. We find out that it helps to enhance the discrimination in classification by integrating other distance based features and/or adding signal preprocessing to the original samples. In this paper, we propose an improved version of the CRC method which uses the Gabor wavelet transformation to preprocess the samples and also adapts the nearest neighbor (NN) features, and hence we call it GNN-CRC. Firstly, Gabor wavelet transformation is applied to minimize the effects from the background in face images and build Gabor features into the input data. Secondly, the distances solved by NN and CRC are fused together to obtain a more discriminative classification. Extensive experiments are conducted to evaluate the proposed method for face recognition with different instantiations. The experimental results illustrate that our method outperforms the naive CRC as well as some other state-of-the-art algorithms.

Key words: face recognition| collaborative representation| Gabor wavelet transformation| nearest neighbor (NN)| image classification

摘要： Collaborative representation-based classification (CRC) is a distance based method, and it obtains the original contributions from all samples to solve the sparse representation coefficient. We find out that it helps to enhance the discrimination in classification by integrating other distance based features and/or adding signal preprocessing to the original samples. In this paper, we propose an improved version of the CRC method which uses the Gabor wavelet transformation to preprocess the samples and also adapts the nearest neighbor (NN) features, and hence we call it GNN-CRC. Firstly, Gabor wavelet transformation is applied to minimize the effects from the background in face images and build Gabor features into the input data. Secondly, the distances solved by NN and CRC are fused together to obtain a more discriminative classification. Extensive experiments are conducted to evaluate the proposed method for face recognition with different instantiations. The experimental results illustrate that our method outperforms the naive CRC as well as some other state-of-the-art algorithms.

关键词: face recognition| collaborative representation| Gabor wavelet transformation| nearest neighbor (NN)| image classification

CLC Number:

TP 391.4

ZHANG Yanghao (张洋豪), ZENG Shaoning (曾少宁), ZENG Wei (曾威), GOU Jianping (苟建平). GNN-CRC: Discriminative Collaborative Representation-Based Classification via Gabor Wavelet Transformation and Nearest Neighbor [J]. Journal of Shanghai Jiao Tong University (Science), 2018, 23(5): 657-665.

References 25

[1]	XU Y, ZHU Q, FAN Z Z, et al. Using the idea ofthe sparse representation to perform coarse-to-fine facerecognition [J]. Information Sciences, 2013, 238(7):138-148.
[2]	ZENG S N, YANG X, GOU J P. Using kernel sparserepresentation to perform coarse-to-fine recognition offace images [J]. Optik-International Journal for Lightand Electron Optics, 2017, 140: 528-535.
[3]	ZHANG L, YANG M, FENG X C. Sparse representationor collaborative representation: Which helps facerecognition? [C]// IEEE International Conference onComputer Vision. [s.l.]: IEEE, 2012: 471-478.
[4]	XU Y, LI Z M, PAN J S, et al. Face recognition basedon fusion of multi-resolution Gabor features [J]. NeuralComputing & Applications, 2013, 23(5): 1251-1256.
[5]	JIA S, HU J, XIE Y, et al. Gabor cube selection basedmultitask joint sparse representation for hyperspectralimage classification [J]. IEEE Transactions on Geoscienceand Remote Sensing, 2016, 54(6): 3174-3187.
[6]	WANG W, WANG R P, SHAN S G, et al. Probabilisticnearest neighbor search for robust classification offace image sets [C]// IEEE International Conferenceand Workshops on Automatic Face and Gesture Recognition.[s.l.]: IEEE, 2015: 1-7.
[7]	KASEMSUMRAN P, AUEPHANWIRIYAKUL S,THEERA-UMPON N. Face recognition using stringgrammar fuzzy K-nearest neighbor [C]// InternationalConference on Knowledge and Smart Technology. [s.l.]:IEEE, 2016: 584-596.
[8]	SERRANO ′A, DE DIEGO I M, CONDE C, et al. Recentadvances in face biometrics with Gabor wavelets:A review [J]. Pattern Recognition Letters, 2010, 31(5):372-381.
[9]	LV X Q, WU J F, LIU W. Face image feature selectionbased on Gabor feature and recursive feature elimination[C]//Sixth International Conference on IntelligentHuman-Machine Systems and Cybernetics. [s.l.]:IEEE, 2014: 266-269.
[10]	ABDULRAHMAN M, GWADABE T R, ABDU F J,et al. Gabor wavelet transform based facial expressionrecognition using PCA and LBP [C]//Signal Processingand Communications Applications Conference.[s.l.]: IEEE, 2014: 2265-2268.
[11]	CHEN X, RAMADGE P J. Collaborative representation,sparsity or nonlinearity: What is key to dictionarybased classification? [C]//IEEE InternationalConference on Acoustics, Speech and Signal Processing.IEEE, 2014: 5227-5231.
[12]	ZENG S N, GOU J P, DENG L M. An antinoise sparserepresentation method for robust face recognition viajoint l1 and l2 regularization [J]. Expert Systems withApplications, 2017, 82: 1-9.
[13]	CAI S J, ZHANG L, ZUO W M, et al. A probabilisticcollaborative representation based approach for patternclassification [C]// Computer Vision and PatternRecognition. [s.l.]: IEEE, 2016: 2950-2959.
[14]	ZENG S N, YANG X, GOU J P. Multiplication fusionof sparse and collaborative representation for robustface recognition [J]. Multimedia Tools & Applications,2016, 76(20): 20889-20907.
[15]	ZENG S N, GOU J P, YANG X. Improvingsparsity of coefficients for robust sparse andcollaborative representation-based image classification[J]. Neural Computing & Applications, 2017.https://doi.org/10.1007/s00521-017-2900-4 (publishedonline).
[16]	XU Y, ZHU Q, CHEN Y, et al. An improvement to thenearest neighbor classifier and face recognition experiments[J]. International Journal of Innovative ComputingInformation and Control, 2013, 9(2): 543-554.
[17]	FENG Q, PAN J S, YAN L. Nearest feature centreclassifier for face recognition [J]. Electronics Letters,2012, 48(18): 1120-1122.
[18]	TANG B, HE H. ENN: Extended nearest neighbormethod for pattern recognition [research frontier][J]. IEEE Computational Intelligence Magazine, 2015,10(3): 52-60.
[19]	WANG Y Z, JHA S, CHAUDHURI K. Analyzingthe robustness of nearest neighbors to adversarialexamples [EB/OL]. (2018-03-11) [2018-05-09].https://arxiv.org/abs/1706.03922.
[20]	WRIGHT J, YANG A Y, GANESH A, et al. Robustface recognition via sparse representation [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2009, 31(2): 210-227.
[21]	PHILLIPS P J, MOON H, RIZVI S A, et al. TheFERET evaluation methodology for face-recognitionalgorithms [J]. IEEE Transactions on Pattern Analysisand Machine Intelligence, 2000, 22(10): 1090-1104.
[22]	GEORGHIADES A S, BELHUMEUR P N,KRIEGMAN D J. From few to many: Illuminationcone models for face recognition under variablelighting and pose [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence, 2002, 23(6):643-660.
[23]	HUANG G B, MATTAR M, BERG T, et al. Labeledfaces in the wild: A database for studying face recognitionin unconstrained environments [C]//Dans Workshopon Faces in Real-Life Images: Detection, Alignment,and Recognition. [s.l.]: Inria, HAL, 2008: 1-15.
[24]	KIM S J, KOH K, LUSTIG M, et al. An interior-pointmethod for large-scale 1-regularized least squares [J].IEEE Journal of Selected Topics in Signal Processing,2007, 1(4): 606-617.
[25]	BECK A, TEBOULLE M. A fast iterative shrinkagethresholdingalgorithm for linear inverse problems [J].SIAM Journal on Imaging Sciences, 2009, 2(1): 183-202.