Skip to main content
SpringerLink
Log in

Learning mean curvature-based regularization to solve the inverse variational problems from noisy data

  • Original Paper
  • Published:
Signal, Image and Video Processing Aims and scope Submit manuscript

Abstract

As an emerging mathematical tool, inverse variational problem approximation (IVPA) has some real applications. Recently, deep learning is used to detect physical phenomena and carry out various tasks, which has achieved promising results. Several works have shown that can learn inverse variational problems (IVPs). In this work, we proposed a new framework to solve mean-curvature-based regularization IVPs from noisy data with different levels based on physical constrained learning and automatic differentiation. Furthermore, traditional variational methods and neural networks-based approaches are integrated, to learn the non-convex and high nonlinear inverse variational problems. And several experiments show that the effectiveness and robustness of our algorithm.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

References

  1. Kobler, E., Effland, A., Kunisch, K., Pock, T.: Total deep variation for linear inverse problems. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 7549–7558 (2020)

  2. Bredies, K., Kunisch, K., Pock, T.: Total generalized variation. SIAM J. Image Sci. 3(3), 492–526 (2010)

    Article  MathSciNet  MATH  Google Scholar 

  3. Li, C., et al. Robust PCL Discovery of Data-Driven Mean-Field Game Systems and Control Problems. Circuits and Syst. I: Regular Papers, IEEE Transactions on PP. 99, 1–14 (2021)

  4. Li, F., Zeng, T.: A universal variational framework for sparsity-based image inpainting. IEEE Trans. Image Process. 23, 4242–4254 (2014)

    Article  MathSciNet  MATH  Google Scholar 

  5. Dobrosotskaya, J.A., Bertozzi, A.: A wavelet-laplace variational technique for image deconvolution and inpainting. IEEE Trans. Image Process. 17, 657–663 (2008)

    Article  MathSciNet  Google Scholar 

  6. Tihonov, A.N.: Solution of incorrectly formulated problems and the regularization method. Soviet Math. 4, 1035–1038 (1963)

    MathSciNet  Google Scholar 

  7. Bae, E., Shi, J., Tai, X.: Graph cuts for curvature based image denoising. IEEE Trans. Image Process. 20, 1199–1210 (2011)

    Article  MathSciNet  MATH  Google Scholar 

  8. Chumchob, N.: Vectorial total variation-based regularization for variational image registration. IEEE Trans. Image Process. 22, 4551–4559 (2013)

    Article  MathSciNet  MATH  Google Scholar 

  9. Gundogdu, E., Constantin, V., Parashar, S., et al. GarNet++: Improving Fast and Accurate Static 3D Cloth Draping by Curvature Loss.[J]. IEEE Trans. pattern. Anal. Mach. Intell. 44(1, 181–195 (2022)

  10. Zhong, Q., Yin, K., Duan, Y.: Image reconstruction by minimizing curvatures on image surface. J. Math. Imaging Vis. 63, 30–55 (2021)

    Article  MathSciNet  MATH  Google Scholar 

  11. Li, C., Yang, Y., Liang, H., Wu, B.: Transfer learning for establishment of recognition of COVID-19 on CT imaging using small-sized training datasets. Knowl. Based Syst. 218, 106849–106849 (2021)

    Article  Google Scholar 

  12. Sun, J., Li, H., Xu, Z. et al.: Deep ADMM-Net for compressive sensing MRI. In: Advances in Neural Information Processing Systems, pp. 10–18 (2016)

  13. Jin, K.H., McCann, M.T., Froustey, E., Unser, M.: Deep convolutional neural network for inverse problems in imaging. IEEE Trans. Image Process. 26(9), 4509–4522 (2017)

    Article  MathSciNet  MATH  Google Scholar 

  14. Csáji, B.C., et al.: Approximation with artificial neural networks. Faculty of Sciences, Etvs Lornd University, Hungary 24(48), 7 (2001)

  15. Mumford, D.: Elastica and computer vision. In: Algebraic Geometry and Its Applications, pp. 491–506. Springer (1994)

  16. Fazel, M., Hindi, H., Boyd, S.P.: Log-det Heuristic for matrix rank minimization with applications to Hankel and Euclidean distance matrices. In: Proceedings of the 2003 American Control Conference, 2003., vol. 3, pp. 2156–2162. IEEE (2003)

  17. Chambolle, A., Pock, T.: Total roto-translational variation. Numer. Math. 142(3), 611–666 (2019)

    Article  MathSciNet  MATH  Google Scholar 

  18. Raissi, M., Perdikaris, P., Karniadakis, G.E.: Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys. 378, 686–707 (2019)

    Article  MathSciNet  MATH  Google Scholar 

  19. Sirignano, J.A., Spiliopoulos, K.: DGM: a deep learning algorithm for solving partial differential equations. J. Comput. Phys. 375, 1339–1364 (2018)

    Article  MathSciNet  MATH  Google Scholar 

  20. Dissanayake, M., Phan-Thien, N.: Neural-network-based approximations for solving partial differential equations. Commun. Numer. Methods Eng. 10(3), 195–201 (1994)

    Article  MATH  Google Scholar 

  21. Long, Z., Lu, Y., Ma, X., Dong, B.: PDE-Net: learning PDEs from Data. In: International Conference on Machine Learning, pp. 3208–3216 (2018)

  22. Meng, X., Babaee, H., Karniadakis, G.E.: Multi-fidelity Bayesian neural networks: algorithms and applications. J. Comput. Phys. 438, 110361 (2021)

    Article  MathSciNet  MATH  Google Scholar 

  23. Cai, S., Wang, Z., Lu, L., Zaki, T.A., Karniadakis, G.E.: DeepM &Mnet: inferring the electroconvection multiphysics fields based on operator approximation by neural networks. J. Comput. Phys. 436, 110296 (2021)

    Article  MathSciNet  MATH  Google Scholar 

  24. El-Fallah, A.I., Ford, G.: Mean curvature evolution and surface area scaling in image filtering. IEEE Trans. Image Process. 6(5), 750–3 (1997)

    Article  Google Scholar 

  25. Tang, M., Shekhar, R., Huang, D.: Mean curvature mapping for detection of corneal shape abnormality. IEEE Trans. Med. Imaging 24, 424–428 (2005)

    Article  Google Scholar 

  26. Kobler, E., Klatzer, T., Hammernik, K., Pock, T.: Variational networks: connecting variational methods and deep learning. In: German conference on pattern recognition, pp. 281–293, Springer (2017)

  27. Adler, J., Öktem, O.: Learned primal-dual reconstruction. IEEE Trans. Med. Imaging 37(6), 1322–1332 (2018)

    Article  Google Scholar 

  28. Effland, A., Kobler, E., Kunisch, K., et al. Variational Networks: An Optimal Control Approach to Early Stopping Variational Methods for Image Restoration[J]. J. Math. Imaging and Vision. 62(2),(2020)

  29. Aggarwal, H.K., Mani, M.P., Jacob, M.: MoDL: model-based deep learning architecture for inverse problems. IEEE Trans. Med. Imaging 38(2), 394–405 (2018)

    Article  Google Scholar 

  30. Ye, J.C., Han, Y., Cha, E.: Deep convolutional framelets: a general deep learning framework for inverse problems. SIAM J. Imag. Sci. 11(2), 991–1048 (2018)

    Article  MathSciNet  MATH  Google Scholar 

  31. Sanghvi, Y., Kalepu, Y., Khankhoje, U.K.: Embedding deep learning in inverse scattering problems. IEEE Trans. Comput. Imaging 6, 46–56 (2019)

    Article  Google Scholar 

  32. Goldstein, T., Osher, S.: The split Bregman method for L1-regularized problems. SIAM J. Imaging Sci. 2, 323–343 (2009)

    Article  MathSciNet  MATH  Google Scholar 

  33. Chambolle, A., Pock, T.: A first-order primal-dual algorithm for convex problems with applications to imaging. J. Math. Imaging Vis. 40, 120–145 (2010)

    Article  MathSciNet  MATH  Google Scholar 

  34. Wang, Y., Yin, W., Zeng, J.: Global convergence of ADMM in nonconvex nonsmooth optimization. J. Sci. Comput. 78, 29–63 (2019)

    Article  MathSciNet  MATH  Google Scholar 

  35. Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. CoRR, vol. arXiv:1412.6980

  36. Zhuang, J., Tang, T., Ding, Y., Tatikonda, S., Dvornek, N., Papademetris, X., Duncan, J.: AdaBelief optimizer: adapting stepsizes by the belief in observed gradients. ArXiv, vol. arXiv:2010.07468 (2020)

  37. Heo, B., Chun, S., Oh, S.J., Han, D., Yun, S., Uh, Y., Ha, J.-W.: Slowing down the weight norm increase in momentum-based optimizers. ArXiv, vol. arXiv:2006.08217 (2020)

  38. Ding, J., Ren, X., Luo, R., Sun, X.: An adaptive and momental bound method for stochastic learning. ArXiv, vol. arXiv:1910.12249 (2019)

  39. Ginsburg, B., Castonguay, P., Hrinchuk, O., Kuchaiev, O., Lavrukhin, V., Leary, R., Li, J., Nguyen, H., Cohen, J.M.: Stochastic gradient methods with layer-wise adaptive moments for training of deep networks. ArXiv, vol. arXiv:1905.11286 (2019)

  40. Lu, L., Meng, X., Mao, Z., Karniadakis, G.: DeepXDE: a deep learning library for solving differential equations. SIAM Rev. 63, 208–228 (2020)

    Article  MathSciNet  MATH  Google Scholar 

  41. Wang, Z., Bovik, A., Sheikh, H., Simoncelli, E.P.: Image quality assessment: from error visibility to structural similarity. IEEE Trans. Image Process. 13, 600–612 (2004)

    Article  Google Scholar 

  42. Hore, A., Ziou, D.: Image quality metrics: PSNR vs. SSIM. In: 2010 20th International Conference on Pattern Recognition, pp. 2366–2369. IEEE (2010)

Download references

Funding

Funding was provided by Natural Science Foundation of China (Grant No. 61731003).

Author information

Author notes

  1. Hongchen Liu and Hongbo Qu contributed equally to this work.

Authors and Affiliations

  1. School of Electrical and Information Engineering, Tianjin University, Tianjin, 300072, China

    Hongchen Liu, Chunping Hou & Yonghong Hou

  2. Tianjin International Engineering Institute, Tianjin University, Tianjin, 300072, China

    Hongbo Qu

Authors
  1. Hongchen Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Chunping Hou
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Hongbo Qu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yonghong Hou
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hongchen Liu.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, H., Hou, C., Qu, H. et al. Learning mean curvature-based regularization to solve the inverse variational problems from noisy data. SIViP 17, 3193–3200 (2023). https://doi.org/10.1007/s11760-023-02544-9

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11760-023-02544-9

Keywords

Search

Navigation