General and Nested Wiberg Minimization: L2 and Maximum Likelihood
Wiberg matrix factorization breaks a matrix Y into low-rank factors U and V by solving for V in closed form given U, linearizing V(U) about U, and iteratively minimizing ||Y − UV(U)||2 with respect to U only. This approach factors the matrix while effectively removing V from the minimization. We generalize the Wiberg approach beyond factorization to minimize an arbitrary function that is nonlinear in each of two sets of variables. In this paper we focus on the case of L 2 minimization and maximum likelihood estimation (MLE), presenting an L 2 Wiberg bundle adjustment algorithm and a Wiberg MLE algorithm for Poisson matrix factorization. We also show that one Wiberg minimization can be nested inside another, effectively removing two of three sets of variables from a minimization. We demonstrate this idea with a nested Wiberg algorithm for L 2 projective bundle adjustment, solving for camera matrices, points, and projective depths.
KeywordsMatrix Factorization Bundle Adjustment Projective Depth Simultaneous Minimization Global Motion Estimate
Unable to display preview. Download preview PDF.
- 1.Wiberg, T.: Computation of principal components when data are missing. In: Second Symposium of Computation Statistics, Berlin, pp. 229–326 (1976)Google Scholar
- 2.Okatani, T., Deguchi, K.: On the Wiberg algorithm for matrix factorization in the presence of missing components. International Journal of Computer Vision 72(3) (May 2007)Google Scholar
- 3.Okatani, T., Yoshida, T., Deguchi, K.: Efficient algorithm for low-rank matrix factorization with missing components and performance comparison of latest algorithms. In: International Conference on Computer Vision, Barcelona, Spain (November 2011)Google Scholar
- 4.Eriksson, A., van den Hengel, A.: Efficient computation of robust low-rank matrix approximations in the presence of missing data using the L 1 norm. In: Computer Vision and Pattern Recognition, San Francisco, CA (June 2010)Google Scholar
- 5.Strelow, D.: General and nested wiberg minimization. In: Computer Vision and Pattern Recognition, Providence, RI (June 2012)Google Scholar
- 6.Ke, Q., Kanade, T.: Robust L 1 norm factorization in the presence of outliers and missing data by alternative convex programming. In: Computer Vision and Pattern Recognition, San Diego, CA (June 2005)Google Scholar
- 9.Poelman, C.: The paraperspective and projective factorization methods for recovering shape and motion. PhD thesis, Carnegie Mellon University (1995)Google Scholar
- 10.Fackler, P.L.: Notes on matrix calculus (September 2005), http://www4.ncsu.edu/~pfackler/MatCalc.pdf (accessed August 29, 2011)
- 14.Forsyth, D.A., Ponce, J.: Computer vision: A modern approach. Prentice-Hall, Upper Saddle River (2003)Google Scholar
- 15.Hartley, R., Zisserman, A.: Multiple view geometry in computer vision, 2nd edn. Cambridge University Press, Cambridge (2003)Google Scholar