Skip to main content

2018 | OriginalPaper | Buchkapitel

A Deep Boltzmann Machine-Based Approach for Robust Image Denoising

verfasst von : Rafael G. Pires, Daniel S. Santos, Gustavo B. Souza, Aparecido N. Marana, Alexandre L. M. Levada, João Paulo Papa

Erschienen in: Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications

Verlag: Springer International Publishing

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

A Deep Boltzmann Machine (DBM) is composed of a stack of learners called Restricted Boltzmann Machines (RBMs), which correspond to a specific kind of stochastic energy-based networks. In this work, a DBM is applied to a robust image denoising by minimizing the contribution of some of its top nodes, called “noise nodes”, which often get excited when noise pixels are present in the given images. After training the DBM with noise and clean images, the detection and deactivation of the noise nodes allow reconstructing images with great quality, eliminating most of their noise. The results obtained from important public image datasets showed the validity of the proposed approach.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Fußnoten
1
Notice the procedure over the noisy images do not consider the activation field already computed for the clean images.
 
2
In this paper, we evaluated \(T\in [0.1, 0.9]\) with steps of 0.1.
 
3
Since we have \(28\times 28\) images concerning all datasets, the visible layer has \(28\,\times \,28 = 784\) nodes. Also, since we are using Bernoulli-based DBM/DBN models, we employed a min-max normalization of the grayscale values of the images’ pixels.
 
4
Since Semeion database images are \(16\,\times \,16\)-sized, we centered them into a \(28\,\times \,28\)-black-squared window in order to have all images used in the experiments with the very same size.
 
Literatur
1.
Zurück zum Zitat Papa, J.P., Fonseca, L.M.G., Carvalho, L.A.S.: Projections onto convex sets through particle swarm optimization and its application for remote sensing image restoration. Pattern Recognit. Lett. 31(13), 1876–1886 (2010)CrossRef Papa, J.P., Fonseca, L.M.G., Carvalho, L.A.S.: Projections onto convex sets through particle swarm optimization and its application for remote sensing image restoration. Pattern Recognit. Lett. 31(13), 1876–1886 (2010)CrossRef
2.
Zurück zum Zitat Pires, R.G., Pereira, D.R., Pereira, L.A.M., Mansano, A.F., Papa, J.P.: Projections onto convex sets parameter estimation through harmony search and its application for image restoration. Nat. Comput. 15(3), 493–502 (2016)MathSciNetCrossRef Pires, R.G., Pereira, D.R., Pereira, L.A.M., Mansano, A.F., Papa, J.P.: Projections onto convex sets parameter estimation through harmony search and its application for image restoration. Nat. Comput. 15(3), 493–502 (2016)MathSciNetCrossRef
3.
Zurück zum Zitat Katsaggelos, A.K.: Digital Image Restoration. Springer, New York (1991)CrossRef Katsaggelos, A.K.: Digital Image Restoration. Springer, New York (1991)CrossRef
4.
Zurück zum Zitat Paik, J.K., Katsaggelos, A.K.: Image restoration using a modified Hopfield network. IEEE Trans. Image Process. 1, 49–63 (1992)CrossRef Paik, J.K., Katsaggelos, A.K.: Image restoration using a modified Hopfield network. IEEE Trans. Image Process. 1, 49–63 (1992)CrossRef
5.
Zurück zum Zitat Sun, Y., Yu, S.Y.: A modified Hopfield neural network used in bilevel image restoration and reconstruction. In: International Symposium on Information Theory Application, vol. 3, pp. 1412–1414 (1992) Sun, Y., Yu, S.Y.: A modified Hopfield neural network used in bilevel image restoration and reconstruction. In: International Symposium on Information Theory Application, vol. 3, pp. 1412–1414 (1992)
6.
Zurück zum Zitat Keyvanrad, M.A., Pezeshki, M., Homayounpour, M.A.: Deep belief networks for image denoising. arXiv preprint arXiv:1312.6158 (2013) Keyvanrad, M.A., Pezeshki, M., Homayounpour, M.A.: Deep belief networks for image denoising. arXiv preprint arXiv:​1312.​6158 (2013)
7.
Zurück zum Zitat Tang, Y., Salakhutdinov, R., Hinton, G.: Robust Boltzmann machines for recognition and denoising. In: IEEE Conference on Computer Vision and Pattern Recognition, Providence, Rhode Island, USA (2012) Tang, Y., Salakhutdinov, R., Hinton, G.: Robust Boltzmann machines for recognition and denoising. In: IEEE Conference on Computer Vision and Pattern Recognition, Providence, Rhode Island, USA (2012)
8.
Zurück zum Zitat Xie, J., Xu, L., Chen, E.: Image denoising and inpainting with deep neural networks. In: Pereira, F., Burges, C.J.C., Bottou, L., Weinberger, K.Q. (eds.) Advances in Neural Information Processing Systems, vol. 25, pp. 341–349. Curran Associates Inc. (2012) Xie, J., Xu, L., Chen, E.: Image denoising and inpainting with deep neural networks. In: Pereira, F., Burges, C.J.C., Bottou, L., Weinberger, K.Q. (eds.) Advances in Neural Information Processing Systems, vol. 25, pp. 341–349. Curran Associates Inc. (2012)
9.
Zurück zum Zitat Tang, Y., Salakhutdinov, R., Hinton, G.E.: Robust Boltzmann machines for recognition and denoising. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2012, pp. 2264–2271. IEEE (2012) Tang, Y., Salakhutdinov, R., Hinton, G.E.: Robust Boltzmann machines for recognition and denoising. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2012, pp. 2264–2271. IEEE (2012)
10.
Zurück zum Zitat Yan, R., Shao, L.: Image blur classification and parameter identification using two-stage deep belief networks. In: 24th British Machine Vision Conference, pp. 1–11 (2013) Yan, R., Shao, L.: Image blur classification and parameter identification using two-stage deep belief networks. In: 24th British Machine Vision Conference, pp. 1–11 (2013)
11.
Zurück zum Zitat Osindero, S., Hinton, G.E., Teh, Y.W.: A fast learning algorithm for deep belief nets. Neural Comput. 18(7), 1527–1554 (2006)MathSciNetCrossRefMATH Osindero, S., Hinton, G.E., Teh, Y.W.: A fast learning algorithm for deep belief nets. Neural Comput. 18(7), 1527–1554 (2006)MathSciNetCrossRefMATH
12.
Zurück zum Zitat Salakhutdinov, R., Hinton, G.E.: Deep Boltzmann machines. In: AISTATS, vol. 1, p. 3 (2009) Salakhutdinov, R., Hinton, G.E.: Deep Boltzmann machines. In: AISTATS, vol. 1, p. 3 (2009)
13.
Zurück zum Zitat Hinton, G.E.: Training products of experts by minimizing contrastive divergence. Neural Comput. 14(8), 1771–1800 (2002)CrossRefMATH Hinton, G.E.: Training products of experts by minimizing contrastive divergence. Neural Comput. 14(8), 1771–1800 (2002)CrossRefMATH
14.
Zurück zum Zitat Salakhutdinov, R., Hinton, G.E.: An efficient learning procedure for deep Boltzmann machines. Neural Comput. 24(8), 1967–2006 (2012)MathSciNetCrossRefMATH Salakhutdinov, R., Hinton, G.E.: An efficient learning procedure for deep Boltzmann machines. Neural Comput. 24(8), 1967–2006 (2012)MathSciNetCrossRefMATH
15.
Zurück zum Zitat Corinna, C., LeCun, Y., Burges, C.J.C.: The MNIST database of handwritten digits (1998) Corinna, C., LeCun, Y., Burges, C.J.C.: The MNIST database of handwritten digits (1998)
16.
Zurück zum Zitat Lichman, M.: UCI machine learning repository (2013) Lichman, M.: UCI machine learning repository (2013)
17.
Zurück zum Zitat Holub, A., Griffin, G., Perona, P.: Caltech-256 object category dataset (2007) Holub, A., Griffin, G., Perona, P.: Caltech-256 object category dataset (2007)
Metadaten
Titel
A Deep Boltzmann Machine-Based Approach for Robust Image Denoising
verfasst von
Rafael G. Pires
Daniel S. Santos
Gustavo B. Souza
Aparecido N. Marana
Alexandre L. M. Levada
João Paulo Papa
Copyright-Jahr
2018
DOI
https://doi.org/10.1007/978-3-319-75193-1_63

Premium Partner