Abstract
In this work we express resource-efficient MAP inference as joint optimization problem w.r.t. (i) messages (i.e. reparametrizations) and (ii) surrogate potentials that are upper bounds for the problem of interest and allow efficient inference. We show that resulting nested optimization task can be solved on trees by a convergent and efficient algorithm, and that its loopy extension also returns convincing MAP solutions in practice. We demonstrate the utility of the method on dense correspondence and image completion problems.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Stricly speaking we do not use a majorizer, but only (less constrained) upper bounds.
- 2.
We focus on problems with at most pairwise cliques, which are most relevant in practice, but everything can be generalized to higher order cliques straightforwardly.
- 3.
The expression \({\mathbf x}_\alpha \setminus x_s\) is shorthand for \(\{ {\mathbf x}'_\alpha : x'_s = x_s \}\). We will also write compactly \(\alpha \ni s\) instead of \(\{ \alpha : s \in \alpha \}\).
- 4.
We prefer the term “resident set” over “particles”, since—in contrast to particle message passing—our method maintains messages also for non-resident states.
- 5.
One can make the algorithm (trivially) convergent e.g. by conditionally updating the primal solution, such that the solution with minimal primal objective so far is always reported.
- 6.
And \(4.7\%\) when using the weaker product label space relaxation [20] with unary potentials being computed on the fly.
References
Baker, S., Scharstein, D., Lewis, J., Roth, S., Black, M.J., Szeliski, R.: A database and evaluation methodology for optical flow. IJCV 92(1), 1–31 (2011)
Besag, J.: On the statistical analysis of dirty pictures. J. R. Stat. Soc. B 48, 259–302 (1986)
Besse, F., Rother, C., Fitzgibbon, A., Kautz, J.: PMBP: patchmatch belief propagation for correspondence field estimation. IJCV 110(1), 2–13 (2014)
Boykov, Y., Veksler, O., Zabih, R.: Fast approximate energy minimization via graph cuts. IEEE Trans. Pattern Anal. Mach. Intell. 23(11), 1222–1239 (2001)
Drory, A., Haubold, C., Avidan, S., Hamprecht, F.A.: Semi-global matching: a principled derivation in terms of message passing. In: Jiang, X., Hornegger, J., Koch, R. (eds.) GCPR 2014. LNCS, vol. 8753, pp. 43–53. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-11752-2_4
Globerson, A., Jaakkola, T.: Fixing max-product: convergent message passing algorithms for MAP LP-relaxations. In: NIPS (2007)
Hazan, T., Shashua, A.: Norm-prodcut belief propagtion: primal-dual message-passing for LP-relaxation and approximate-inference. IEEE Trans. Inform. Theory 56(12), 6294–6316 (2010)
Hirschmüller, H.: Accurate and efficient stereo processing by semi-global matching and mutual information. In: Proceedings of the CVPR, pp. 807–814 (2005)
Ihler, A., McAllester, D.: Particle belief propagation. In: AISTATS, pp. 256–263 (2009)
Kolmogorov, V.: Convergent tree-reweighted message passing for energy minimization. IEEE Trans. Pattern Anal. Mach. Intell. 28(10), 1568–1583 (2006)
Komodakis, N., Tziritas, G.: Image completion using efficient belief propagation via priority scheduling and dynamic pruning. IEEE Trans. Image Proc. 16(11), 2649–2661 (2007)
Kothapa, R., Pacheco, J., Sudderth, E.: Max-product particle belief propagation. Technical report, Master’s project report, Brown University, Department of Computer Science (2011)
Lempitsky, V., Rother, C., Roth, S., Blake, A.: Fusion moves for Markov random field optimization. IEEE Trans. Pattern Anal. Mach. Intell. 32(8), 1392–1405 (2010)
Li, Y., Min, D., Brown, M.S., Do, M.N., Lu, J.: SPM-BP: sped-up patchmatch belief propagation for continuous MRFs. In: Proceedings of the ICCV, pp. 4006–4014 (2015)
Noorshams, N., Wainwright, M.J.: Stochastic belief propagation: a low-complexity alternative to the sum-product algorithm. IEEE Trans. Inform. Theory 59(4), 1981–2000 (2013)
Pacheco, J., Sudderth, E.: Proteins, particles, and pseudo-max-marginals: a submodular approach. In: Proceedings of the ICML, pp. 2200–2208 (2015)
Peng, J., Hazan, T., McAllester, D., Urtasun, R.: Convex max-product algorithms for continuous MRFs with applications to protein folding. In: Proceedings of the ICML (2011)
Roig, G., Boix, X., Nijs, R.D., Ramos, S., Kuhnlenz, K., Gool, L.V.: Active map inference in CRFs for efficient semantic segmentation. In: Proceedings of ICCV, pp. 2312–2319 (2013)
Scharstein, D., Szeliski, R.: High-accuracy stereo depth maps using structured light. In: Proceedings of the CVPR, pp. 195–202 (2003)
Shekhovtsov, A., Kovtun, I., Hlaváč, V.: Efficient MRF deformation model for non-rigid image matching. CVIU 112(1), 91–99 (2008)
Shekhovtsov, A., Reinbacher, C., Graber, G., Pock, T.: Solving dense image matching in real-time using discrete-continuous optimization. arXiv preprint arXiv:1601.06274 (2016)
Sontag, D., Globerson, A., Jaakkola, T.: Introduction to dual decomposition for inference. In: Optimization for Machine Learning. MIT Press (2011)
Sontag, D., Jaakkola, T.: Tree block coordinate descent for MAP in graphical models. J. Mach. Learn. Res. (2009)
Trinh, H., McAllester, D.: Unsupervised learning of stereo vision with monocular cues. In: Proceedings of the BMVC, pp. 72–81 (2009)
Wainwright, M.J., Jaakkola, T.S., Willsky, A.S.: MAP estimation via agreement on trees: message-passing and linear programming. IEEE Trans. Inf. Theory 51(11), 3697–3717 (2005)
Werner, T.: A linear programming approach to max-sum problem: a review. IEEE Trans. Pattern Anal. Mach. Intell. 29(7) (2007)
Zach, C.: A principled approach for coarse-to-fine MAP inference. In: Proceedings of the CVPR, pp. 1330–1337 (2014)
Zach, C.: Limited-memory belief propagation via nested optimization (2017). Supplementary material https://sites.google.com/site/christophermzach/home/pdf/lmbp_supp.pdf
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this paper
Cite this paper
Zach, C. (2018). Limited-Memory Belief Propagation via Nested Optimization. In: Pelillo, M., Hancock, E. (eds) Energy Minimization Methods in Computer Vision and Pattern Recognition. EMMCVPR 2017. Lecture Notes in Computer Science(), vol 10746. Springer, Cham. https://doi.org/10.1007/978-3-319-78199-0_34
Download citation
DOI: https://doi.org/10.1007/978-3-319-78199-0_34
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-78198-3
Online ISBN: 978-3-319-78199-0
eBook Packages: Computer ScienceComputer Science (R0)