Abstract
An important class of computational techniques to solve inverse problems in image processing relies on a variational approach: the optimal output is obtained by finding a minimizer of an energy function or “model” composed of two terms, the data-fidelity term, and the regularization term. Much research has focused on models where both terms are convex, which leads to convex optimization problems. However, there is evidence that non-convex regularization can improve significantly the output quality for images characterized by some sparsity property. This fostered recent research toward the investigation of optimization problems with non-convex terms. Non-convex models are notoriously difficult to handle as classical optimization algorithms can get trapped at unwanted local minimizers. To avoid the intrinsic difficulties related to non-convex optimization, the convex non-convex (CNC) strategy has been proposed, which allows the use of non-convex regularization while maintaining convexity of the total cost function. This work focuses on a general class of parameterized non-convex sparsity-inducing separable and non-separable regularizers and their associated CNC variational models. Convexity conditions for the total cost functions and related theoretical properties are discussed, together with suitable algorithms for their minimization based on a general forward-backward (FB) splitting strategy. Experiments on the two classes of considered separable and non-separable CNC variational models show their superior performance than the purely convex counterparts when applied to the discrete inverse problem of restoring sparsity-characterized images corrupted by blur and noise.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bauschke, H.H., Combettes, P.L.: Convex Analysis and Monotone Operator Theory in Hilbert Spaces. Springer, New York (2011)
Bayram, I.: On the convergence of the iterative shrinkage/thresholding algorithm with a weakly convex penalty. IEEE Trans. Signal Process. 64(6), 1597–1608 (2016)
Beck, A., Teboulle, M.: A fast iterative shrinkage-thresholding algorithm for linear inverse problems. SIAM J. Imag. Sci. 2(1), 183–202 (2009)
Becker, S., Combettes, P.L.: An algorithm for splitting parallel sums of linearly composed monotone operators, with applications to signal recovery. J. Nonlinear Convex Anal. 15(1), 137–159 (2014)
Bello Cruz, J.Y.: On proximal subgradient splitting method for minimizing the sum of two nonsmooth convex functions. Set-Valued Var. Anal 25, 245–263 (2017)
Blake, A., Zisserman, A.: Visual Reconstruction. MIT Press, Cambridge, MA (1987)
Bruckstein, A., Donoho, D., Elad, M.: From sparse solutions of systems of equations to sparse modeling of signals and images. SIAM Rev. 51(1), 34–81 (2009)
Burger, M., Papafitsoros, K., Papoutsellis, E., Schönlieb, C.B.: Infimal convolution regularisation functionals of BV and Lp spaces. J. Math. Imaging Vis. 55(3), 343–369 (2016)
Cai, G., Selesnick, I.W., Wang, S., Dai, W., Zhu, Z.: Sparsity enhanced signal decomposition via generalized minimax-concave penalty for gearbox fault diagnosis. J. Sound Vib. 432, 213–234 (2018)
Candés, E.J., Wakin, M.B., Boyd, S.: Enhancing sparsity by reweighted l1 minimization. J. Fourier Anal. Appl.14(5), 877–905 (2008)
Carlsson, M.: On convexification/optimization of functionals including an l2-misfit term. arXiv preprint arXiv:1609.09378 (2016)
Castella, M., Pesquet, J.C.: Optimization of a Geman-McClure like criterion for sparse signal deconvolution. In: IEEE International Workshop on Computational Advances Multi-sensor Adaptive Processing, pp. 309–312 (2015)
Chambolle, A., Lions, P.L.: Image recovery via total variation minimization and related problems. Numerische Mathematik 76, 167–188 (1997)
Chan, R., Lanza, A., Morigi, S., Sgallari, F.: Convex non-convex image segmentation. Numerische Mathematik 138(3), 635–680 (2017)
Chartrand, R.: Shrinkage mappings and their induced penalty functions. In: International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 1026–1029 (2014)
Chen, P.Y., Selesnick, I.W.: Group-sparse signal denoising: non-convex regularization, convex optimization. IEEE Trans. Signal Proc. 62, 3464–3478 (2014)
Chouzenoux, E., Jezierska, A., Pesquet, J., Talbot, H.: A majorize-minimize subspace approach for l2–l0 image regularization. SIAM J. Imag. Sci. 6(1), 563–591 (2013)
Ding, Y., Selesnick, I.W.: Artifact-free wavelet denoising: nonconvex sparse regularization, convex optimization. IEEE Signal Process. Lett. 22(9), 1364–1368 (2015)
Du, H., Liu, Y.: Minmax-concave total variation denoising. Signal Image Video Process. 12(6), 1027–1034 (2018)
Geiger, D., Girosi, F.: Parallel and deterministic algorithms from MRF’s: surface reconstruction. IEEE Trans. Pattern Anal. Mach. Intell. 13(5), 410–412 (1991)
Geman, S., Geman, D.: Stochastic relaxation, Gibbs distribution, and the Bayesian restoration of images. IEEE PAMI 6(6), 721–741 (1984)
Hansen, P.C.: Rank-Deficient and Discrete Ill-Posed Problems. SIAM, Philadelphia (1997)
Hartman, P.: On functions representable as a difference of convex functions. Pac. J. Math. 9(3), 707–713 (1959)
Hastie, T., Tibshirani, R., Wainwright, M.: Statistical Learning with Sparsity: The Lasso and Generalizations. CRC Press, Boca Raton (2015)
Huska, M., Lanza, A., Morigi, S., Sgallari, F.: Convex non-convex segmentation of scalar fields over arbitrary triangulated surfaces. J. Comput. Appl. Math. 349, 438–451 (2019a)
Huska, M., Lanza, A., Morigi, S., Selesnick, I.W.: A convex-nonconvex variational method for the additive decomposition of functions on surfaces. Inverse Problems 35, 124008–124041 (2019b)
Jensen, J.B., Nielsen, M.: A simple genetic algorithm applied to discontinuous regularization. In: Proceedings IEEE workshop on NNSP, Copenhagen (1992)
Lanza, A., Morigi, S., Sgallari, F.: Convex image denoising via non-convex regularization. Scale Space Variat. Methods Comput. Vis. 9087, 666–677 (2015)
Lanza, A., Morigi, S., Sgallari, F.: Convex image denoising via nonconvex regularization with parameter selection. J. Math. Imaging Vis. 56(2), 195–220 (2016a)
Lanza, A., Morigi, S., Sgallari, F.: Constrained TVp-ℓ2 model for image restoration. J. Sci. Comput. 68, 64–91 (2016b)
Lanza, A., Morigi, S., Selesnick, I.W., Sgallari, F.: Nonconvex nonsmooth optimization via convex-nonconvex majorization minimization. Numerische Mathematik 136(2), 343–381 (2017)
Lanza, A., Morigi, S., Sgallari, F.: Automatic parameter selection based on residual whiteness for convex non-convex variational restoration. In: Mathematical Methods in Image Processing and and Inverse Problems (eds) Tai XC, Wei S, Liu H. Springer, Singapore, 360, (2021). https://doi.org/10.1007/978-981-16-2701-9
Lanza, A., Morigi, S., Selesnick, I.W., Sgallari, F.: Sparsity-inducing nonconvex nonseparable regularization for convex image processing. SIAM J. Imag. Sci. 12(2), 1099–1134 (2019)
Lanza, A., Pragliola, M., Sgallari, F.: Residual whiteness principle for parameter-free image restoration. Electron. Trans. Numer. Anal. 53, 329–351 (2020)
Lefkimmiatis, S., Ward, J., Unser, M.: Hessian Schatten-Norm regularization for linear inverse problems. IEEE Trans. Image Process. 22, 1873–1888 (2013)
Malek-Mohammadi, M., Rojas, C.R., Wahlberg, B.: A class of nonconvex penalties preserving overall convexity in optimization based mean filtering. IEEE Trans. Signal Process. 64(24), 6650–6664 (2016)
Nikolova, M.: Estimation of binary images by minimizing convex criteria. Proc. IEEE Int. Conf. Image Process. 2, 108–112 (1998)
Nikolova, M.: Energy minimization methods. In: Scherzer, O. (ed.) Handbook of Mathematical Methods in Imaging, Chapter 5, pp. 138–186. Springer, Berlin (2011)
Nikolova, M., Ng, M.K., Tam, C.P.: Fast nonconvex nonsmooth minimization methods for image restoration and reconstruction. IEEE Trans. Image Process. 19(12), 3073–3088 (2010)
Parekh, A., Selesnick, I.W.: Convex denoising using non-convex tight frame regularization. IEEE Signal Process. Lett. 22(10), 1786–1790 (2015)
Parekh, A., Selesnick, I.W.: Enhanced low-rank matrix approximation. IEEE Signal Process. Lett. 23(4), 493–497 (2016)
Park, T.W., Burrus, C.S.: Digital Filter Design. Wiley, New York (1987)
Portilla, J., Mancera, L.: L0-based sparse approximation: two alternative methods and some applications. In: Proceedings of SPIE, San Diego, vol. 6701 (Wavelets XII) (2007)
Rudin, L.I., Osher, S., Fatemi, E.: Nonlinear total variation based noise removal algorithms. Physics D 60(1–4), 259–268 (1992)
Selesnick, I.W.: Sparse regularization via convex analysis. IEEE Trans. Signal Process. 65(17), 4481–4494 (2017a)
Selesnick, I.W.: Total variation denoising via the Moreau envelope. IEEE Signal Process. Lett. 24(2), 216–220 (2017b)
Selesnick, I.W., Bayram, I.: Sparse signal estimation by maximally sparse convex optimization. IEEE Trans. Signal Proc. 62(5), 1078–1092 (2014)
Selesnick, I.W., Parekh, A., Bayram, I.: Convex 1-D total variation denoising with non-convex regularization. IEEE Signal Process. Lett. 22, 141–144 (2015)
Selesnick, I.W., Lanza, A., Morigi, S., Sgallari, F.: Non-convex total variation regularization for convex denoising of signals. J. Math. Imag. Vis. 62, 825–841 (2020)
Setzer, S., Steidl, G., Teuber, T.: Infimal convolution regularizations with discrete l1-type functionals. Commun. Math. Sci. 9(3), 797–827 (2011)
Shen, L., Xu, Y., Zeng, X.: Wavelet inpainting with the l0 sparse regularization. J. Appl. Comp. Harm. Anal. 41(1), 26–53 (2016)
Sidky, E.Y., Chartrand, R., Boone, J.M., Pan, X.: Constrained TpV–minimization for enhanced exploitation of gradient sparsity: application to CT image reconstruction. IEEE J. Trans. Eng. Health Med. 2, 1–18 (2014)
Soubies, E., Blanc-Féraud, L., Aubert, G.: A continuous exact L0 penalty (CEL0) for least squares regularized problem. SIAM J. Imag. Sci.8(3), 1607–1639 (2015)
Tipping, M.E.: Sparse Bayesian learning and the relevance vector machine. J. Mach. Learn. Res. 1, 211–244 (2001)
Tuy, H.: DC optimization: theory, methods and algorithms. In: Handbook of Global Optimization, pp. 149–216. Springer, Boston, (1995)
Wang, S., Selesnick, I.W., Cai, G., Ding, B., Chen, X.: Synthesis versus analysis priors via generalized minimax-concave penalty for sparsity-assisted machinery fault diagnosis. Mech. Syst. Signal Process. 127, 202–233 (2019)
Wipf, D.P., Rao, B.D., Nagarajan, S.: “Latent variable Bayesian models for promoting sparsity. In: IEEE Trans. Inf. Theory 57(9), 6236–6255 (2011)
Yuille, A.L., Rangarajan, A.: The concave-convex procedure. Neural Comput. 15(4), 915–936 (2003)
Zhang, C.-H.: Nearly unbiased variable selection under minimax concave penalty. Ann. Stat. 38(2), 894–942 (2010)
Zou, J., Shen, M., Zhang, Y., Li, H., Liu, G., Ding, S.: Total variation denoising with non-convex regularizers. IEEE Access 7, 4422–4431 (2019)
Acknowledgements
This research was supported in part by the National Group for Scientific Computation (GNCS-INDAM), Research Projects 2019/2020.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 Springer Nature Switzerland AG
About this entry
Cite this entry
Lanza, A., Morigi, S., Selesnick, I.W., Sgallari, F. (2023). Convex Non-convex Variational Models. In: Chen, K., Schönlieb, CB., Tai, XC., Younes, L. (eds) Handbook of Mathematical Models and Algorithms in Computer Vision and Imaging. Springer, Cham. https://doi.org/10.1007/978-3-030-98661-2_61
Download citation
DOI: https://doi.org/10.1007/978-3-030-98661-2_61
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-98660-5
Online ISBN: 978-3-030-98661-2
eBook Packages: Mathematics and StatisticsReference Module Computer Science and Engineering