Skip to main content
SpringerLink
Log in

Nonnegative low rank tensor approximations with multidimensional image applications

  • Published:
Numerische Mathematik Aims and scope Submit manuscript

Abstract

The main aim of this paper is to develop a new algorithm for computing a nonnegative low rank tensor approximation for nonnegative tensors that arise in many multidimensional imaging applications. Nonnegativity is one of the important properties, as each pixel value refers to a nonzero light intensity in image data acquisitions. Our approach is different from classical nonnegative tensor factorization (NTF), which requires each factorized matrix, and/or tensor, to be nonnegative. In this paper, we determine a nonnegative low Tucker rank tensor to approximate a given nonnegative tensor. We propose an alternating projections algorithm for computing such a nonnegative low rank tensor approximation, which is referred to as NLRT. The convergence of the proposed manifold projection method is established. The experimental results for synthetic data and multidimensional images are presented to demonstrate that the performance of NLRT is better than the state-of-the-art NTF methods.

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
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10

Similar content being viewed by others

Notes

  1. \(\{ \mathcal{Z}^{(k)} \}\) falls in \(\texttt{U}\) means there exists a positive integer \(k_0\) such that \(\{ \mathcal{Z}^{(k)} \}_{k \ge k_0} \subset \texttt{U}\).

  2. To avoid making the entries negative, we first simulate noise with a standard normal distribution and then set the negative noisy value to 0. The SNR in dB is defined as \(\text {SNR}_\text {dB} = 20\log _{10}\frac{\Vert \mathcal {X}_\text {groundtruth}\Vert _F}{\Vert \text {Noise}\Vert _F}\).

  3. Defined as \( \frac{\Vert \mathcal {X}_\text {estimated}-\mathcal {X}_\text {groundtruth}\Vert _F}{\Vert \mathcal {X}_\text {groundtruth}\Vert _F}. \)

  4. Videos are available at http://trace.eas.asu.edu/yuv/ and https://sites.google.com/site/jamiezeminzhang/publications.

  5. Data available at http://www.ehu.eus/ccwintco/index.php?title=Hyperspectral_Remote_Sensing_Scenes.

  6. Data available at https://engineering.purdue.edu/~biehl/MultiSpec/hyperspectral.html.

  7. Data available at https://www.cs.rochester.edu/~jliu/code/TensorCompletion.zip.

  8. Data available at https://aviris.jpl.nasa.gov/data/free_data.html.

  9. Data available at http://openremotesensing.net/knowledgebase/hyperspectral-video/.

  10. https://en.wikipedia.org/wiki/Peak_signal-to-noise_ratio

  11. Data available at https://engineering.purdue.edu/\(\sim \)biehl/MultiSpec/hyperspectral.html.

References

  1. Attouch, H., Bolte, J., Redont, P., Soubeyran, A.: Proximal alternating minimization and projection methods for nonconvex problems: an approach based on the kurdyka-łojasiewicz inequality. Math. Oper. Res. 35(2), 438–457 (2010)

    Article  MathSciNet  MATH  Google Scholar 

  2. Balasubramaniam, T., Nayak, R., Yuen, C.: Efficient nonnegative tensor factorization via saturating coordinate descent. ACM Trans. Knowl. Discov. Data (TKDD) 14(4), 1–28 (2020)

    Article  Google Scholar 

  3. Bauschke, H.H., Luke, D.R., Phan, H.M., Wang, X.: Restricted normal cones and the method of alternating projections: applications. Set Value Var. Anal. 21(3), 475–501 (2013)

    Article  MathSciNet  MATH  Google Scholar 

  4. Bauschke, H.H., Luke, D.R., Phan, H.M., Wang, X.: Restricted normal cones and the method of alternating projections: theory. Set Value Var. Anal. 21(3), 431–473 (2013)

    Article  MathSciNet  MATH  Google Scholar 

  5. Cichocki, A., Phan, A.H.: Fast local algorithms for large scale nonnegative matrix and tensor factorizations. IEICE Trans. Fundam. Electron. Commun. Comput. Sci. 92(3), 708–721 (2009)

    Article  Google Scholar 

  6. Cichocki, A., Zdunek, R., Amari, S.i.: Hierarchical ALS algorithms for nonnegative matrix and 3D tensor factorization. In: International conference on independent component analysis and signal separation, pp. 169–176. Springer (2007)

  7. Clarke, F., Vinter, R.: Regularity properties of optimal controls. SIAM J. Control Optim. 28(4), 980–997 (1990)

    Article  MathSciNet  MATH  Google Scholar 

  8. De Lathauwer, L., De Moor, B., Vandewalle, J.: A multilinear singular value decomposition. SIAM J. Matrix Anal. Appl. 21(4), 1253–1278 (2000)

    Article  MathSciNet  MATH  Google Scholar 

  9. De Lathauwer, L., De Moor, B., Vandewalle, J.: On the best rank-1 and rank-(r 1, r 2,..., rn) approximation of higher-order tensors. SIAM J. Matrix Anal. Appl. 21(4), 1324–1342 (2000)

    Article  MathSciNet  MATH  Google Scholar 

  10. Drusvyatskiy, D., Ioffe, A., Lewis, A.: Alternating projections and coupling slope. arXiv preprint arXiv:1401.7569 pp. 1–17 (2014)

  11. Golub, G.H., Van Loan, C.F.: Matrix Computations, vol. 3. JHU Press, Baltimore (2012)

    MATH  Google Scholar 

  12. Kim, Y.D., Choi, S.: Nonnegative Tucker decomposition. In: 2007 IEEE conference on computer vision and pattern recognition, pp. 1–8. IEEE (2007)

  13. Kolda, T.G., Bader, B.W.: Tensor decompositions and applications. SIAM Rev. 51(3), 455–500 (2009)

    Article  MathSciNet  MATH  Google Scholar 

  14. Kroonenberg, P.M.: Applied Multiway Data Analysis, vol. 702. Wiley, New York (2008)

    Book  MATH  Google Scholar 

  15. Lewis, A.S., Luke, D.R., Malick, J.: Local linear convergence for alternating and averaged nonconvex projections. Found. Comput. Math. 9(4), 485–513 (2009)

    Article  MathSciNet  MATH  Google Scholar 

  16. Lewis, A.S., Luke, D.R., Malick, J.: Local linear convergence for alternating and averaged nonconvex projections. Found. Comput. Math. 9(4), 485–513 (2009)

    Article  MathSciNet  MATH  Google Scholar 

  17. Lewis, A.S., Malick, J.: Alternating projections on manifolds. Math. Oper. Res. 33(1), 216–234 (2008)

    Article  MathSciNet  MATH  Google Scholar 

  18. Li, G., Pong, T.K.: Douglas-rachford splitting for nonconvex optimization with application to nonconvex feasibility problems. Math. Program. 159(1–2), 371–401 (2016)

    Article  MathSciNet  MATH  Google Scholar 

  19. Li, X., Ng, M.K., Cong, G., Ye, Y., Wu, Q.: MR-NTD: manifold regularization nonnegative tucker decomposition for tensor data dimension reduction and representation. IEEE Trans. Neural Netw. Learn. Syst. 28(8), 1787–1800 (2016)

    Article  MathSciNet  Google Scholar 

  20. Noll, D., Rondepierre, A.: On local convergence of the method of alternating projections. Found. Comput. Math. 16(2), 425–455 (2016)

    Article  MathSciNet  MATH  Google Scholar 

  21. Pan, J., Ng, M.K., Liu, Y., Zhang, X., Yan, H.: Orthogonal nonnegative tucker decomposition. arXiv preprint arXiv:1912.06836 (2019)

  22. Rockafellar, R.T., Wets, R.J.B.: Variational Analysis, vol. 317. Springer, Cham (2009)

    MATH  Google Scholar 

  23. Shin, K., Sael, L., Kang, U.: Fully scalable methods for distributed tensor factorization. IEEE Trans. Knowl. Data Eng. 29(1), 100–113 (2016)

    Article  Google Scholar 

  24. Sidiropoulos, N.D., De Lathauwer, L., Fu, X., Huang, K., Papalexakis, E.E., Faloutsos, C.: Tensor decomposition for signal processing and machine learning. IEEE Trans. Signal Process. 65(13), 3551–3582 (2017)

    Article  MathSciNet  MATH  Google Scholar 

  25. Song, G.J., Ng, M.K.: Nonnegative low rank matrix approximation for nonnegative matrices. Appl. Math. Lett. (2020). https://doi.org/10.1016/j.aml.2020.106300

    Article  MathSciNet  MATH  Google Scholar 

  26. Tucker, L.R.: Some mathematical notes on three-mode factor analysis. Psychometrika 31(3), 279–311 (1966)

    Article  MathSciNet  Google Scholar 

  27. Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P.: Image quality assessment: from error visibility to structural similarity. IEEE Trans. Image Process. 13(4), 600–612 (2004)

    Article  Google Scholar 

  28. Welling, M., Weber, M.: Positive tensor factorization. Pattern Recognit. Lett. 22(12), 1255–1261 (2001)

    Article  MATH  Google Scholar 

  29. Xu, Y., Hao, R., Yin, W., Su, Z.: Parallel matrix factorization for low-rank tensor completion. Inverse Probl. Imaging 9(2), 601–624 (2015)

    Article  MathSciNet  MATH  Google Scholar 

  30. Xu, Y., Yin, W.: A block coordinate descent method for regularized multiconvex optimization with applications to nonnegative tensor factorization and completion. SIAM J. Imaging Sci. 6(3), 1758–1789 (2013)

    Article  MathSciNet  MATH  Google Scholar 

  31. Zhang, Y., Zhou, G., Zhao, Q., Cichocki, A., Wang, X.: Fast nonnegative tensor factorization based on accelerated proximal gradient and low-rank approximation. Neurocomputing 198, 148–154 (2016)

    Article  Google Scholar 

  32. Zhou, G., Cichocki, A., Xie, S.: Fast nonnegative matrix/tensor factorization based on low-rank approximation. IEEE Trans. Signal Process. 60(6), 2928–2940 (2012)

    Article  MathSciNet  MATH  Google Scholar 

Download references

Acknowledgements

T.-X. Jiang’s research is supported in part by the National Natural Science Foundation of China under Grant 12001446, the Natural Science Foundation of Sichuan, China under Grant 2022NSFSC1798, and the Fundamental Research Funds for the Central Universities under Grants JBK2202049 and JBK2102001. M. K. Ng’s research is supported in part by Hong Kong Research Grant Council GRF 12300218, 12300519, 17201020, 17300021, C1013-21GF, C7004-21GF and Joint NSFC-RGC N-HKU76921. G.-J. Song’s research is supported in part by the National Natural Science Foundation of China under Grant 12171369 and Key NSF of Shandong Province under Grant ZR2020KA008.

Author information

Authors and Affiliations

  1. School of Computing and Artificial Intelligence, Southwestern University of Finance and Economics, Chengdu, People’s Republic of China

    Tai-Xiang Jiang

  2. Kash Institute of Electronics and Information Industry, Kashi, People’s Republic of China

    Tai-Xiang Jiang

  3. Department of Mathematics, The University of Hong Kong, Pokfulam, Hong Kong, People’s Republic of China

    Michael K. Ng & Junjun Pan

  4. School of Mathematics and Information Sciences, Weifang University, Weifang, 261061, People’s Republic of China

    Guang-Jing Song

Authors
  1. Tai-Xiang Jiang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Michael K. Ng
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Junjun Pan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Guang-Jing Song
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Guang-Jing Song.

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

Jiang, TX., Ng, M.K., Pan, J. et al. Nonnegative low rank tensor approximations with multidimensional image applications. Numer. Math. 153, 141–170 (2023). https://doi.org/10.1007/s00211-022-01328-6

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00211-022-01328-6

Mathematics Subject Classification

Search

Navigation