×

Orthogonal incremental non-negative matrix factorization algorithm and its application in image classification. (English) Zbl 1449.68062

Summary: To improve the sparseness of the base matrix in incremental non-negative matrix factorization, we in this paper present a new method, orthogonal incremental non-negative matrix factorization algorithm (OINMF), which combines the orthogonality constraint with incremental learning. OINMF adopts batch update in the process of incremental learning, and its iterative formulae are obtained using the gradient on the Stiefel manifold. The experiments on image classification show that the proposed method achieves much better sparseness and orthogonality, while retaining time efficiency of incremental learning.

MSC:

68T05 Learning and adaptive systems in artificial intelligence
15A23 Factorization of matrices
62H30 Classification and discrimination; cluster analysis (statistical aspects)
62H35 Image analysis in multivariate analysis
68W25 Approximation algorithms
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Bucak, SS; Gunsel, B., Incremental subspace learning via non-negative matrix factorization, Pattern Recogn, 42, 5, 788-797 (2009) · Zbl 1178.68395 · doi:10.1016/j.patcog.2008.09.002
[2] Cai, D.; He, X.; Han, J.; Huang, TS, Graph regularized nonnegative matrix factorization for data representation, IEEE Trans Pattern Anal Mach Intell, 33, 8, 1548-1560 (2011) · doi:10.1109/TPAMI.2010.231
[3] Cichocki A, Zdunek R, Amari S (2006) New algorithms for non-negative matrix factorization in applications to blind source separation. In: Acoustics, speech and signal processing, 2006. ICASSP 2006 Proceedings. 2006 IEEE international conference on, IEEE, vol 5, pp V-V · Zbl 1178.94057
[4] Cui, Z.; Cao, Z.; Yang, J.; Feng, J.; Ren, H., Target recognition in synthetic aperture radar images via non-negative matrix factorisation, IET Radar Sonar Navigat, 9, 9, 1376-1385 (2015) · doi:10.1049/iet-rsn.2014.0407
[5] Dang S, Cui Z, Cao Z, Liu Y, Min R (2017) Sar target recognition via incremental nonnegative matrix factorization with lp sparse constraint. In: Radar conference, pp 0530-0534
[6] Edelman, A.; Arias, TA; Smith, ST, The geometry of algorithms with orthogonality constraints, SIAM J Matrix Anal Appl, 20, 303-353 (1998) · Zbl 0928.65050 · doi:10.1137/S0895479895290954
[7] Guillamet D, Vitrià J (2002) Non-negative matrix factorization for face recognition. In: Topics in artificial intelligence, Springer, New York, pp 336-344 · Zbl 1028.68642
[8] Hirzer M, Beleznai C, Roth PM, Bischof H (2011) Person re-identification by descriptive and discriminative classification. In: Proceedings of Scandinavian conference on image analysis (SCIA)
[9] Hoyer, PO, Non-negative matrix factorization with sparseness constraints, J Mach Learn Res, 5, Nov, 1457-1469 (2004) · Zbl 1222.68218
[10] King B, Févotte C, Smaragdis P (2012) Optimal cost function and magnitude power for nmf-based speech separation and music interpolation. In: Machine learning for signal processing (MLSP), 2012 IEEE international workshop on, IEEE, pp 1-6
[11] Lee, DD; Seung, HS, Learning the parts of objects by non-negative matrix factorization, Nature, 401, 6755, 788-791 (1999) · Zbl 1369.68285 · doi:10.1038/44565
[12] Lee DD, Seung HS (2001) Algorithms for non-negative matrix factorization. In: Leen TK, Dietterich TG, Tresp V (eds) Advances in neural information processing systems 13, MIT Press, pp 556-562, http://papers.nips.cc/paper/1861-algorithms-for-non-negative-matrix-factorization.pdf
[13] Liu F, Liu M, Zhou T, Qiao Y, Yang J (2016) Incremental robust nonnegative matrix factorization for object tracking. In: International conference on neural information processing, Springer, New York, pp 611-619
[14] Liu, H.; Wu, Z.; Li, X.; Cai, D.; Huang, TS, Constrained nonnegative matrix factorization for image representation, IEEE Trans Pattern Anal Mach Intell, 34, 7, 1299-1311 (2012) · doi:10.1109/TPAMI.2011.217
[15] Li, L.; Zhang, YJ, Linear projection-based non-negative matrix factoriztion, ACTA Autom Sin, 36, 1, 23-39 (2010) · doi:10.3724/SP.J.1004.2010.00023
[16] Stiefel, E., Richtungsfelder und fernparallelismus in n-dimensionalen mannigfaltigkeiten, Comment Math Helvet, 8, 1, 305-353 (1935) · Zbl 0014.41601 · doi:10.1007/BF01199559
[17] Wang, CYJ; Sun, F., Incremental learning algorithm based on graph regularized non-negative matrix factorization with sparseness constraints, J Comput Appl, 37, 4, 1071-1074 (2017)
[18] Wang D, Lu H (2011) Incremental orthogonal projective non-negative matrix factorization and its applications. In: 2011 18th IEEE international conference on image processing, pp 2077-2080. 10.1109/ICIP.2011.6115890
[19] Wang, D.; Lu, H., On-line learning parts-based representation via incremental orthogonal projective non-negative matrix factorization, Sig Process, 93, 6, 1608-1623 (2013) · doi:10.1016/j.sigpro.2012.07.015
[20] Wang, W.; Cai, J., Incremental learning algorithm of non-negative matrix factorization with sparseness constraints, Comput Sci, 41, 8, 241-244 (2014)
[21] Yang, Z.; Yuan, Z.; Laaksonen, J., Projective non-negative matrix factorization with applications to facial image processing, Int J Pattern Recogn Artif Intell, 21, 8, 1353-1362 (2007) · doi:10.1142/S0218001407005983
[22] Yoo, J.; Choi, S.; Fyfe, C.; Kim, D.; Lee, SY; Yin, H., Orthogonal nonnegative matrix factorization: Multiplicative updates on stiefel manifolds, Intelligent data engineering and automated learning—IDEAL 2008, 140-147 (2008), Berlin: Springer, Berlin
[23] Yu, ZZ; Liu, YH; Li, B.; Pang, SC; Jia, CC, Incremental graph regulated nonnegative matrix factorization for face recognition, J Appl Math, 1, 1-10 (2014) · Zbl 1442.68217
[24] Yuan, Z.; Oja, E.; Kalviainen, H.; Parkkinen, J.; Kaarna, A., Projective nonnegative matrix factorization for image compression and feature extraction, Image Anal, 333-342 (2005), Heidelberg: Springer, Heidelberg
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.