Learning Multi-view Correspondences via Subspace-Based Temporal Coincidences

  • Christian Conrad
  • Rudolf Mester
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7944)


In this work we present an approach to automatically learn pixel correspondences between pairs of cameras. We build on the method of Temporal Coincidence Analysis (TCA) and extend it from the pure temporal (i.e. single-pixel) to the spatiotemporal domain. Our approach is based on learning a statistical model for local spatiotemporal image patches, determining rare, and expressive events from this model, and matching these events across multiple views. Accumulating multi-image coincidences of such events over time allows to learn the desired geometric and photometric relations. The presented method also works for strongly different viewpoints and camera settings, including substantial rotation, and translation. The only assumption that is made is that the relative orientation of pairs of cameras may be arbitrary, but fixed, and that the observed scene shows visual activity. We show that the proposed method outperforms the single pixel approach to TCA both in terms of learning speed and accuracy.


Image Patch Epipolar Geometry Event Match Temporal Coincidence Spatiotemporal Domain 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Conrad, C., Guevara, A., Mester, R.: Learning multi-view correspondences from temporal coincidences. In: CVPR Workshops, pp. 9–16. IEEE (2011)Google Scholar
  2. 2.
    Lowe, D.: Distinctive image features from scale-invariant keypoints. IJCV (2004)Google Scholar
  3. 3.
    Rosten, E., Drummond, T.W.: Machine learning for high-speed corner detection. In: Leonardis, A., Bischof, H., Pinz, A. (eds.) ECCV 2006, Part I. LNCS, vol. 3951, pp. 430–443. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  4. 4.
    Leutenegger, S., Chli, M., Siegwart, R.: Brisk: Binary robust invariant scalable keypoints. In: ICCV, pp. 2548–2555 (2011)Google Scholar
  5. 5.
    Laptev, I.: On space-time interest points. IJCV 64(2), 107–123 (2005)MathSciNetCrossRefGoogle Scholar
  6. 6.
    Harris, C., Stephens, M.: A combined corner and edge detector. In: Alvey Vision Conference, Manchester, UK, vol. 15, p. 50 (1988)Google Scholar
  7. 7.
    Förstner, W., Gülch, E.: A fast operator for detection and precise location of distinct points, corners and centres of circular features. In: ISPRS, pp. 281–305 (1987)Google Scholar
  8. 8.
    Szlávik, Z., Sziranyi, T., Havasi, L., Benedek, C.: Optimizing of searching co-motion point-pairs for statistical camera calibration. In: ICIP, vol. 2, pp. 1178–1181 (2005)Google Scholar
  9. 9.
    Szlávik, Z., Havasi, L., Szirányi, T.: Geometrical scene analysis using co-motion statistics. In: Blanc-Talon, J., Philips, W., Popescu, D., Scheunders, P. (eds.) ACIVS 2007. LNCS, vol. 4678, pp. 968–979. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  10. 10.
    Szlávik, Z., Szirányi, T., Havasi, L.: Video camera registration using accumulated co-motion maps. ISPRS JPRS 61(5), 298–306 (2007)CrossRefGoogle Scholar
  11. 11.
    Ermis, E., Saligrama, V., Jodoin, P., Konrad, J.: Abnormal behavior detection and behavior matching for networked cameras. In: ICDSC, pp. 1–10. IEEE (2008)Google Scholar
  12. 12.
    Zhang, Z., Deriche, R., Faugeras, O., Luong, Q.: A robust technique for matching two uncalibrated images through the recovery of the unknown epipolar geometry. AI 78(1-2), 87–119 (1995)Google Scholar
  13. 13.
    Wexler, Y., Fitzgibbon, A., Zisserman, A.: Learning epipolar geometry from image sequences. In: Proc. CVPR, vol. 2, p. 209 (2003)Google Scholar
  14. 14.
    Triggs, B.: Joint feature distributions for image correspondence. In: ICCV (2001)Google Scholar
  15. 15.
    Duda, R.O., Hart, P.E., Stork, D.H.: Pattern Classification, 2nd edn. Wiley (2000)Google Scholar
  16. 16.
    Moghaddam, B., Pentland, A.: Probabilistic visual learning for object detection. In: ICCV, p. 786 (1995)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Christian Conrad
    • 1
  • Rudolf Mester
    • 2
    • 1
  1. 1.Visual Sensorics and Information Processing Lab (VSI) Computer Science Dept.Goethe UniversityFrankfurtGermany
  2. 2.Computer Vision Laboratory, Electr. Eng. Dept. (ISY)Linköping UniversitySweden

Personalised recommendations