A BAYESIAN APPROACH TO 3D OBJECT RECOGNITION USING LINEAR COMBINATION OF 2D VIEWS

Vasileios Zografos, Bernard F. Buxton

2008

Abstract

We introduce Bayes priors into a recent pixel-based, linear combination of views object recognition technique. Novel views of an object are synthesized and matched to the target scene image using numerical optimisation. Experiments on a real-image, public database with the use of two different optimisation methods indicate that the priors effectively regularize the error surface and lead to good performance in both cases. Further exploration of the parameter space has been carried out using Markov Chain Monte Carlo sampling.

References

  1. Bishop, C. M. (1995). Neural Networks for Pattern Recognition. Oxford University Press.
  2. Dias, M. B. and Buxton, B. F. (2005). Implicit, view invariant, linear flexible shape modelling. Pattern Recognition Letters, 26(4):433-447.
  3. Gelman, A., Carlin, J. B., Stern, H. S., and Rubin, D. B. (1995). Bayesian Data Analysis. Chapman and Hall, London, 2nd edition.
  4. Koufakis, I. and Buxton, B. F. (1998). Very low bit-rate face video compression using linear combination of 2dfaceviews and principal components analysis. Image and Vision Computing, 17:1031-1051.
  5. Nelder, J. A. and Mead, R. (1965). A simplex method for function minimization. Computer Journal, 7:308- 313.
  6. Nene, S. A., Nayar, S. K., and Murase, H. (1996). Columbia Object Image Library (COIL-20). Technical Report CUCS-006-96, Department of computer science, Columbia University, New York, N.Y. 10027.
  7. Shashua, A. (1995). Algebraic functions for recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 17(8):779-789.
  8. Shewchuk, J. R. (2002). Delaunay refinement algorithms for triangular mesh generation. Computational Geometry: Theory and Applications, 22:21-74.
  9. Storn, R. and Price, K. V. (1997). Differential evolution - a simple and efficient heuristic for global optimization overcontinuous spaces. Journal of Global Optimization, 11(4):341-359.
  10. Ullman, S. and Basri, R. (1991). Recognition by linear combinations of models. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(10):992-1006.
  11. Zografos, V. and Buxton, B. F. (2007). Pose-invariant 3d object recognition using linear combination of 2d views and evolutionary optimisation. ICCTA, pages 645-649.
Download


Paper Citation


in Harvard Style

Zografos V. and F. Buxton B. (2008). A BAYESIAN APPROACH TO 3D OBJECT RECOGNITION USING LINEAR COMBINATION OF 2D VIEWS . In Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2008) ISBN 978-989-8111-21-0, pages 295-298. DOI: 10.5220/0001081902950298


in Bibtex Style

@conference{visapp08,
author={Vasileios Zografos and Bernard F. Buxton},
title={A BAYESIAN APPROACH TO 3D OBJECT RECOGNITION USING LINEAR COMBINATION OF 2D VIEWS},
booktitle={Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2008)},
year={2008},
pages={295-298},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001081902950298},
isbn={978-989-8111-21-0},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2008)
TI - A BAYESIAN APPROACH TO 3D OBJECT RECOGNITION USING LINEAR COMBINATION OF 2D VIEWS
SN - 978-989-8111-21-0
AU - Zografos V.
AU - F. Buxton B.
PY - 2008
SP - 295
EP - 298
DO - 10.5220/0001081902950298