Making metric learning algorithms invariant to transformations using a projection metric on Grassmann manifolds

  • Zahra Goudarzi
  • Peyman AdibiEmail author
  • Rolf-Rainer Grigat
  • Mohammad Saeid Ehsani
Original Article


The requirement for suitable ways to measure the distance or similarity between data is omnipresent in machine learning, pattern recognition and data mining, but extracting such good metrics for particular problems is in general challenging. This has led to the emergence of metric learning ideas, which intend to automatically learn a distance function tuned to a specific task. In many tasks and data types, there are natural transformations to which the classification result should be invariant or insensitive. This demand and its implications are essential in many machine learning applications, and insensitivity to image transformations was in the first place achieved by using invariant feature vectors. In this paper, a new representation model on Grassmann manifolds for data points and a novel method for learning a Mahalanobis metric which uses the geodesic distance on Grassmann manifolds are proposed. In fact, we use an appropriate geodesic distance metric on the Grassmann manifolds, called projection metric, for measuring primary similarities between the new representations of the data points. This makes learning of the Mahalanobis metric invariant to similarity transforms and intensity changes, and therefore improve the performance. Experiments on face and handwritten digit datasets demonstrate that our proposed method yields performance improvements in a state-of-the-art metric learning algorithm.


Metric learning Grassmann manifold Mahalanobis metric 



  1. 1.
    Absil P-A, Mahony R, Sepulchre R (2009) Optimization algorithms on matrix manifolds. Princeton University Press, PrincetonzbMATHGoogle Scholar
  2. 2.
    Bar-Hillel A, Hertz T, Shental N, Weinshall D (2005) Learning a Mahalanobis metric from equivalence constraints. J Mach Learn Res 6:937–965MathSciNetzbMATHGoogle Scholar
  3. 3.
    Bi Y, Fan B, Wu F (2015) Beyond Mahalanobis metric: Cayley-Klein metric learning. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 2339–2347Google Scholar
  4. 4.
    Candemir S, Borovikov E, Santosh KC, Antani S, Thoma G (2015) Rsilc: rotation-and scale-invariant, line-based color-aware descriptor. Image Vis Comput 42:1–12CrossRefGoogle Scholar
  5. 5.
    Crammer K, Singer Y (2001) On the algorithmic implementation of multiclass kernel-based vector machines. J Mach Learn Res 2(Dec):265–292zbMATHGoogle Scholar
  6. 6.
    Davis JV, Kulis B, Jain P, Sra S, Dhillon IS (2007) Information-theoretic metric learning. In: Proceedings of the 24th international conference on machine learning, ACM, pp 209–216.Google Scholar
  7. 7.
    Domeniconi C, Gunopulos D (2002) Adaptive nearest neighbor classification using support vector machines. In: Advances in neural information processing systems, pp 665–672Google Scholar
  8. 8.
    Duda RO, Hart PE, Stork DG (2012) Pattern classification. Wiley, New YorkzbMATHGoogle Scholar
  9. 9.
    Edelman A, Arias TA, Smith ST (1998) The geometry of algorithms with orthogonality constraints. SIAM J Matrix Anal Appl 20(2):303–353MathSciNetCrossRefzbMATHGoogle Scholar
  10. 10.
    Goldberger J, Hinton GE, Roweis ST, Salakhutdinov RR (2005) Neighbourhood components analysis. In: Advances in neural information processing systems, pp 513–520Google Scholar
  11. 11.
    Guillaumin M, Verbeek J, Schmid C (2009) Is that you? Metric learning approaches for face identification. In: 2009 IEEE 12th international conference on computer Vision, pp 498–505. IEEEGoogle Scholar
  12. 12.
    Halmos PR (2012) A Hilbert space problem book, vol 19. Springer Science & Business Media, New YorkzbMATHGoogle Scholar
  13. 13.
    Hamm J, Lee DD (2008) Grassmann discriminant analysis: a unifying view on subspace-based learning. In: Proceedings of the 25th international conference on machine learning, pp 376–383. ACMGoogle Scholar
  14. 14.
    Harandi M, Sanderson C, Shen C, Lovell BC (2013) Dictionary learning and sparse coding on Grassmann manifolds: An extrinsic solution. In: Proceedings of the IEEE international conference on computer vision, pp 3120–3127Google Scholar
  15. 15.
    Harandi MT, Salzmann M, Jayasumana S, Hartley R, Li H (2014) Expanding the family of Grassmannian kernels: an embedding perspective. In: European conference on computer vision, Springer, pp 408–423.Google Scholar
  16. 16.
    Helmke U, Hüper K, Trumpf J (2007) Newton’s method on Grassmann manifolds. arXiv preprint arXiv:0709.2205
  17. 17.
    Hochuli AG, Oliveira LS, Britto AS Jr, Sabourin R (2018) Handwritten digit segmentation: is it still necessary? Pattern Recognit 78:1–11CrossRefGoogle Scholar
  18. 18.
    Hoi SCH, Liu W, Chang S-F (2010) Semi-supervised distance metric learning for collaborative image retrieval and clustering. ACM Trans Multimed Comput Commun Appl (TOMM) 6(3):18Google Scholar
  19. 19.
    Huang Z, Wang R, Shan S, Chen X (2015) Projection metric learning on Grassmann manifold with application to video based face recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 140–149Google Scholar
  20. 20.
    Koestinger M, Hirzer M, Wohlhart P, Roth PM, Bischof H (2012) Large scale metric learning from equivalence constraints. In: 2012 IEEE conference on computer vision and pattern recognition (CVPR) , pp 2288–2295. IEEEGoogle Scholar
  21. 21.
    Le H (1991) On geodesics in Euclidean shape spaces. J Lond Math Soc 2(2):360–372MathSciNetCrossRefzbMATHGoogle Scholar
  22. 22.
    Lim D, Lanckriet G (2014) Efficient learning of Mahalanobis metrics for ranking. In: Proceedings of the 31st international conference on machine learning (ICML-14), pp 1980–1988Google Scholar
  23. 23.
    Patil S, Talbar S (2012) Content based image retrieval using various distance metrics. In: Data engineering and management, Springer, pp 154–161Google Scholar
  24. 24.
    Peng J, Heisterkamp DR, Dai HK (2002) Adaptive kernel metric nearest neighbor classification. In: Proceedings of 16th international conference on pattern recognition, 2002, vol 3, pp 33–36. IEEEGoogle Scholar
  25. 25.
    Santosh KC (2011) Character recognition based on dtw-radon. In: 2011 international conference on document analysis and recognition (ICDAR), pp 264–268. IEEEGoogle Scholar
  26. 26.
    Santosh KC, Aafaque A, Antani S, Thoma GR (2017) Line segment-based stitched multipanel figure separation for effective biomedical cbir. Int J Pattern Recognit Artif Intell 31(06):1757003CrossRefGoogle Scholar
  27. 27.
    Santosh KC, Lamiroy B, Wendling L (2014) Integrating vocabulary clustering with spatial relations for symbol recognition. Int J Doc Anal Recognit (IJDAR) 17(1):61–78CrossRefGoogle Scholar
  28. 28.
    Santosh KC, Roy PP (2018) Arrow detection in biomedical images using sequential classifier. Int J Mach Learn Cybern 9(6):993–1006CrossRefGoogle Scholar
  29. 29.
    Santosh KC, Wendling L, Antani S, Thoma GR (2016) Overlaid arrow detection for labeling regions of interest in biomedical images. IEEE Intell Syst 31(3):66–75CrossRefGoogle Scholar
  30. 30.
    Sethi S, Rohila VK, Agarwal P (2018) Hand written and natural scene character recognition. Hand 5(05):908–911Google Scholar
  31. 31.
    Shen C, Kim J, Wang L, Hengel AVD (2012) Positive semidefinite metric learning using boosting-like algorithms. J Mach Learn Res 13:1007–1036MathSciNetzbMATHGoogle Scholar
  32. 32.
    Srivastava A, Klassen E (2004) Bayesian and geometric subspace tracking. Adv Appl Probab 36(01):43–56MathSciNetCrossRefzbMATHGoogle Scholar
  33. 33.
    Vajda S, Santosh KC (2016) A fast k-nearest neighbor classifier using unsupervised clustering. In: International conference on recent trends in image processing and pattern recognition, Springer, pp 185–193.Google Scholar
  34. 34.
    Wang D, Tan X (2018) Robust distance metric learning via bayesian inference. IEEE Trans Image Process 27(3):1542–1553MathSciNetCrossRefGoogle Scholar
  35. 35.
    Wang J, Kalousis A, Woznica A (2012) Parametric local metric learning for nearest neighbor classification. In: Advances in neural information processing systems, pp 1601–1609Google Scholar
  36. 36.
    Weinberger KQ, Saul LK (2009) Distance metric learning for large margin nearest neighbor classification. J Mach Learn Res 10(Feb):207–244zbMATHGoogle Scholar
  37. 37.
    Xiang S, Nie F, Zhang C (2008) Learning a Mahalanobis distance metric for data clustering and classification. Pattern Recognit 41(12):3600–3612CrossRefzbMATHGoogle Scholar
  38. 38.
    Ying Y, Li P (2012) Distance metric learning with eigenvalue optimization. J Mach Learn Res 13:1–26MathSciNetzbMATHGoogle Scholar
  39. 39.
    Zadeh P, Hosseini R, Sra S (2016) Geometric mean metric learning. In: International conference on machine learning, pp 2464–2471Google Scholar
  40. 40.
    Zuo W, Wang F, Zhang D, Lin L, Huang Y, Meng D, Zhang L (2015) Iterated support vector machines for distance metric learning. arXiv preprint arXiv:1502.00363

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  • Zahra Goudarzi
    • 1
  • Peyman Adibi
    • 1
    Email author
  • Rolf-Rainer Grigat
    • 2
  • Mohammad Saeid Ehsani
    • 1
  1. 1.Department of Artificial Intelligence, Computer Engineering FacultyUniversity of IsfahanIsfahanIran
  2. 2.Institute of Vision SystemsHamburg University of TechnologyHamburgGermany

Personalised recommendations