Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Whiteness-based bilevel learning of regularization parameters in imaging (2403.07026v1)

Published 10 Mar 2024 in math.OC, cs.LG, and eess.IV

Abstract: We consider an unsupervised bilevel optimization strategy for learning regularization parameters in the context of imaging inverse problems in the presence of additive white Gaussian noise. Compared to supervised and semi-supervised metrics relying either on the prior knowledge of reference data and/or on some (partial) knowledge on the noise statistics, the proposed approach optimizes the whiteness of the residual between the observed data and the observation model with no need of ground-truth data.We validate the approach on standard Total Variation-regularized image deconvolution problems which show that the proposed quality metric provides estimates close to the mean-square error oracle and to discrepancy-based principles.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (20)
  1. E. J. Candes, J. Romberg, and T. Tao, “Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information,” IEEE Trans. Inform. Theory, vol. 52, no. 2, 2006.
  2. L. I. Rudin, S. Osher, and E. Fatemi, “Nonlinear total variation based noise removal algorithms,” Physica D: Nonlinear Phen., vol. 60, no. 1, 1992.
  3. L. Calatroni, A. Lanza, M. Pragliola, and F. Sgallari, “Adaptive parameter selection for weighted-TV image reconstruction problems,” Journal of Physics: Conference Series, vol. 1476, no. 1, p. 012003, 2020.
  4. M. Pragliola, L. Calatroni, A. Lanza, and F. Sgallari, “On and beyond total variation regularization in imaging: The role of space variance,” SIAM Rev., vol. 65, no. 3, pp. 601–685, 2023.
  5. P. C. Hansen, “Analysis of discrete ill-posed problems by means of the L-curve,” SIAM Rev., vol. 34, no. 4, pp. 561–580, 1992.
  6. V. A. Morozov, “On the solution of functional equations by the method of regularization,” Doklady Mathematics, vol. 7, pp. 414–417, 1966.
  7. J. Pesquet, A. Benazza-Benyahia, and C. Chaux, “A SURE approach for digital signal/image deconvolution problems,” IEEE Trans. Signal Process., vol. 57, no. 12, 2009.
  8. E. Haber and L. Tenorio, “Learning regularization functionals: a supervised training approach,” Inverse Probl., vol. 19, no. 3, pp. 611–626, 2003.
  9. K. G. G. Samuel and M. F. Tappen, “Learning optimized MAP estimates in continuously-valued MRF models,” in 2009 CVPR, 2009.
  10. F. Pedregosa, “Hyperparameter optimization with approximate gradient,” in ICML, vol. 48.   New York, USA: PMLR, 2016, pp. 737–746.
  11. K. Kunisch and T. Pock, “A bilevel optimization approach for parameter learning in variational models,” SIAM J. on Imaging Sci., vol. 6, no. 2, 2013.
  12. J. C. De los Reyes, C.-B. Schönlieb, and T. Valkonen, “Bilevel parameter learning for higher-order total variation regularisation models,” J. of Math. Imaging Vis., vol. 57, no. 1, 2017.
  13. C. Crockett and J. A. Fessler, “Bilevel methods for image reconstruction,” Found. Trends Signal Process., vol. 15, no. 2-3, 2022.
  14. J. Fehrenbach, M. Nikolova, G. Steidl, and P. Weiss, “Bilevel image denoising using gaussianity tests,” in SSVM.   Cham: Springer International Publishing, 2015, pp. 117–128.
  15. A. Lanza, M. Pragliola, and F. Sgallari, “Residual whiteness principle for parameter-free image restoration,” Electron. Trans. Numer. Anal., vol. 53, pp. 329–351, 2020.
  16. M. Pragliola, L. Calatroni, A. Lanza, and F. Sgallari, “ADMM-based residual whiteness principle for automatic parameter selection in single image super-resolution problems,” J. Math. Imaging Vis., vol. 65, no. 1, pp. 99–123, 2023.
  17. F. Bevilacqua, A. Lanza, M. Pragliola, and F. Sgallari, “Whiteness-based parameter selection for Poisson data in variational image processing,” Appl. Math. Model., vol. 117, pp. 197–218, 2023.
  18. A. Chambolle, “An algorithm for total variation minimization and applications,” J. Math. Imaging Vis., vol. 20, no. 1, pp. 89–97, 2004.
  19. D. R. Martin, C. C. Fowlkes, D. Tal, and J. Malik, “A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics,” ICCV 2001, vol. 2, pp. 416–423 vol.2, 2001.
  20. A. Floquet, S. Dutta, E. Soubies, D.-H. Pham, and D. Kouame, “Automatic tuning of denoising algorithms parameters without ground truth,” IEEE Signal Process. Lett., vol. 31, pp. 381–385, 2024.

Summary

We haven't generated a summary for this paper yet.