Abstract
This chapter proposes to utilize sparse structure for visual information sensing and understanding. In detail, concentered on the fundamental theory of compressive sensing, we will discuss the problem of low-rank structure learning (LRSL) from sparse outliers. Different from traditional approaches, which directly utilize convex norms to measure the sparseness, our method introduces more reasonable non-convex measurements to enhance the sparsity in both the intrinsic low-rank structure and the sparse corruptions. Although the proposed optimization is no longer convex, it still can be effectively solved by a majorization–minimization (MM)-type algorithm. From the theoretic perspective, we have proved that the MM-type algorithm can converge to a stationary point after successive iterations. The proposed model is applied to solve a number of computer vision and information processing tasks, e.g., face image enhancement, object tracking, and time series clustering.
Parts of this chapter are reproduced from [1] with permission number 3410110407533 \(@\) IEEE.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
In cases, the weighting matrices may cause complex numbers due to the inverse operation. In such condition, we use the approximating matrices \(\mathbf {W}_Y=\mathbf {U}(\Sigma +\delta \mathbf {I}_m)^{-1/2}\mathbf {U}^T\) and \(\mathbf {W}_Z=\mathbf {V}(\Sigma +\delta \mathbf {I}_m)^{-1/2}\mathbf {V}^T\) in LHR.
- 2.
The optimization for \(p\)HR is very similar by changing the weight matrices.
- 3.
We only report the result of \(p=1/3\) here since in the previous numerical simulation, \(p\)HR (\(p=1/3\)) achieves higher recovery accuracy than \(p\)HR with \(P=2/3\).
- 4.
For example, in the industrial category of drug manufactures, it is not possible to get the historical data of CIPILA.LTD from [28] which is the only the interface for us to get the stock prices in USA.
References
Hu S, Wang J (2003) Absolute exponential stability of a class of continuous-time recurrent neural networks. IEEE Trans Neural Netw 14(1):35–45
Goldberg AB, Zhu X, Recht B, Xu J-M, Nowak RD (2010) Transduction with matrix completion: three birds with one stone. In: Advances in Neural Information Processing Systems, pp 757–765
Liu G, Lin Z, Yan S, Sun J, Yu Y, Ma Y (2013) Robust recovery of subspace structures by low-rank representation. IEEE Trans Pattern Anal Mach Intell 35(1):171–184
Hu S, Wang J (2001) Quadratic stabilizability of a new class of linear systems with structural independent time-varying uncertainty. Automatica 37(1):51–59. Available http://www.sciencedirect.com/science/article/pii/S0005109800001229
Deng Y, Li Y, Qian Y, Ji X, Dai Q (2014) Visual words assignment via information-theoretic manifold embedding. IEEE Trans Cybern
Deng Y, Liu Y, Dai Q, Zhang Z, Wang Y (2012) Noisy depth maps fusion for multiview stereo via matrix completion. IEEE J Sel Top Sig Process 6(5):566–582
Deng Y, Dai Q, Zhang Z (2011) Graph laplace for occluded face completion and recognition. IEEE Trans Image Process 20(8):2329–2338
Deng Y, Dai Q, Wang R, Zhang Z (2012) Commute time guided transformation for feature extraction. Comput Vis Image Underst 116(4):473–483
Donoho DL (2006) Compressed sensing. IEEE Trans Inf Theor 52(4):1289–1306
Recht B, Fazel M, Parrilo PA (2010) Guaranteed minimum-rank solutions of linear matrix equations via nuclear norm minimization. SIAM Rev 52(3):471–501
Deng Y, Dai Q, Liu R, Zhang Z, Hu S (2013) Low-rank structure learning via nonconvex heuristic recovery. IEEE Trans Neural Netw Learn Syst 24(3):383–396
Candès EJ, Li X, Ma Y, Wright J (2011) Robust principal component analysis? J ACM 58(3): 11
Chandrasekaran V, Sanghavi S, Parrilo PA, Willsky AS (2011) Rank-sparsity incoherence for matrix decomposition. SIAM J Optim 21(2):572–596
Hsu D, Kakade SM, Zhang T (2010) Robust matrix decomposition with outliers. arXiv:1011.1518
Yang J, Yu K, Gong Y, Huang T (2009) Linear spatial pyramid matching using sparse coding for image classification. In: Computer vision and pattern recognition. IEEE Conference on CVPR 2009. IEEE 2009, pp 1794–1801
Fazel M (2002) Matrix rank minimization with applications. Ph.D. dissertation, PhD thesis, Stanford University
Candes EJ, Wakin MB, Boyd SP (2008) Enhancing sparsity by reweighted ? 1 minimization. J Fourier Anal Appl 14(5–6):877–905
Zou H, Hastie T (2005) Regularization and variable selection via the elastic net. J Roy Stat Soc: Series B (Stat Methodol) 67(2):301–320
Foo C-S, Do CB, Ng AY (2009) A majorization-minimization algorithm for (multiple) hyperparameter learning. In: Proceedings of the 26th annual international conference on machine learning. ACM, pp 321–328
Mohan K, Fazel M (2010) Reweighted nuclear norm minimization with application to system identification. In: American control conference (ACC). IEEE, pp 2953–2959
Hunter D, Li R (2005) Variable selection using mm algorithms. Ann Stat 33(4):1617
Lange K (1995) A gradient algorithm locally equivalent to the em algorithm. J Roy Stat Soc Series B (Methodol) 425–437
Lin Z, Chen M, Ma Y (2011) The augmented lagrange multiplier method for exact recovery of corrupted low-rank matrices. Technical report arXiv:1009.5055v2
Benedek C, Sziranyi T (2008) Bayesian foreground and shadow detection in uncertain frame rate surveillance videos. IEEE Trans Image Process 17(4):608–621
Fischler MA, Bolles RC (1981) Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Commun ACM 24:381–395. Available http://doi.acm.org/10.1145/358669.358692
Vidal R, Ma Y, Sastry S (2003) Generalized principal component analysis (gpca). In: Proceedings of 2003 IEEE computer society conference on computer vision and pattern recognition, vol 1, pp I-621– I-628
Shi J, Malik J (2000) Normalized cuts and image segmentation. IEEE Trans Pattern Anal Mach Intell 22(8):888–905
Yahoo!finacial, http://finance.yahoo.com
Google finacial. Available http://www.google.com.hk/finance?q=
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2015 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Deng, Y. (2015). Sparse Structure for Visual Information Sensing: Theory and Algorithms. In: High-Dimensional and Low-Quality Visual Information Processing. Springer Theses. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-44526-6_2
Download citation
DOI: https://doi.org/10.1007/978-3-662-44526-6_2
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-44525-9
Online ISBN: 978-3-662-44526-6
eBook Packages: EngineeringEngineering (R0)