cvpr cvpr2013 cvpr2013-466 cvpr2013-466-reference knowledge-graph by maker-knowledge-mining
Source: pdf
Author: Brian Potetz, Mohammadreza Hajiarbabi
Abstract: For problems over continuous random variables, MRFs with large cliques pose a challenge in probabilistic inference. Difficulties in performing optimization efficiently have limited the probabilistic models explored in computer vision and other fields. One inference technique that handles large cliques well is Expectation Propagation. EP offers run times independent of clique size, which instead depend only on the rank, or intrinsic dimensionality, of potentials. This property would be highly advantageous in computer vision. Unfortunately, for grid-shaped models common in vision, traditional Gaussian EP requires quadratic space and cubic time in the number of pixels. Here, we propose a variation of EP that exploits regularities in natural scene statistics to achieve run times that are linear in both number of pixels and clique size. We test these methods on shape from shading, and we demonstrate strong performance not only for Lambertian surfaces, but also on arbitrary surface reflectance and lighting arrangements, which requires highly non-Gaussian potentials. Finally, we use large, non-local cliques to exploit cast shadow, which is traditionally ignored in shape from shading.
[1] A. H. Ahmed and A. A. Farag. A new formulation for shape from shading for non-lambertian surfaces. In CVPR, 2006. 7
[2] N. Alon and R. Yuster. Solving linear systems through nested dissection. In IEEE 51st Annual Symposium on Foundations of Computer Science, pages 225–234, 2010. 3
[3] J. T. Barron and J. Malik. High-frequency shape and albedo from shading using natural image statistics. In CVPR, 2011. 4
[4] P. N. Belhumeur. A Bayesian approach to binocular steropsis. International Journal of Computer Vision, 19(3):237–
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17] 260, 1996. 5 A. Birlutiu and T. Heskes. Expectation propagation for rating players in sports competitions. In KDD, 2007. 7 A. Ecker and A. D. Jepson. Polynomial shape from shading. In CVPR, pages 145–152, June 2010. 4, 5, 7 D. A. Forsyth. Variable-source shading analysis. Int. J. of Computer Vision, 91(3):280–302, Feb. 2011. 4, 7 J. Huang, A. B. Lee, and D. Mumford. Statistics of range images. In CVPR, pages 1324–133 1, 2000. 2, 3, 6 N. Khan, L. Tram, and M. Tappen. Training many-parameter shape-from-shading models using a surface database. In 3DIM 2009 Workshop at ICCV, pages 1433–1440, 2009. 7 P. Kohli, L. Ladicky, and P. H. S. Torr. Robust higher order potentials for enforcing label consistency. International Journal of Computer Vision, 82(3):302–324, 2009. 1, 3 V. Kolmogorov and R. Zabih. Multi-camera scene reconstruction via graph cuts. In ECCV, 2002. 3 P. Kr ¨ahenb u¨hl and V. Koltun. Efficient inference in fully connected CRFs with gaussian edge potentials. In NIPS, pages 109–1 17. 2011. 1 K. Lee and C. Kuo. Shape from shading with a linear triangular element surface model. IEEE Trans. Pattern Anal. Mach. Intell., 15(8):815–822, 1993. 7 T. P. Minka. Expectation propagation for approximate bayesian inference. In UAI 2001, page 362, 2001 . 1, 2, 4, 7 T. P. Minka. A family of algorithms for approximate Bayesian inference. PhD thesis, Massachusetts Institute of Technology, January 2001 . 1, 2, 5 T. P. Minka and Y. Qi. Tree-structured approximations by expectation propagation. In NIPS, 2004. 2 B. Potetz. Efficient belief propagation for vision using linear
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25] constraint nodes. In CVPR. 2007. 1, 3, 4, 5, 6, 7 B. Potetz and T. S. Lee. Scaling laws in natural scenes and the inference of 3D shape. In NIPS. 2006. 3, 6 E. Prados and O. Faugeras. Perspective shape from shading and viscosity solutions. In ICCV, page 826, 2003. 4, 7 S. Roth and M. J. Black. Fields of experts: A framework for learning image priors. In CVPR, page 860, 2005. 2, 3, 5 D. L. Ruderman. Origins of scaling in natural images. Vision Research, 37:3385–3398, 1997. 3 M. A. J. van Gerven, B. Cseke, F. P. de Lange, and T. Heskes. Efficient Bayesian multivariate fMRI analysis using a sparsifying spatio-temporal prior. NeuroImage, 50(1): 150–161, 2010. 2, 3 O. Vogel, M. Breuß, T. Leichtweis, and J. Weickert. Fast shape from shading for phong-type surfaces. In Proceedings of the Second International Conference on Scale Space and Variational Methods in Computer Vision, 2009. 7 O. J. Woodford, P. H. S. Torr, I. D. Reid, and A. W. Fitzgibbon. Global stereo reconstruction under second order smoothness priors. IEEE Trans. Pattern Anal. Mach. Intell., 31(12):21 15–2128, 2009. 1, 3, 5 J. S. Yedidia, W. T. Freeman, and Y. Weiss. Constructing free-energy approximations and generalized belief propagation algorithms. IEEE Transactions on Information Theory, 51(7):2282–2312, 2005. 5 111666778199