HOP-MAP: Efficient Message Passing with High Order Potentials

There is a growing interest in building probabilistic models with high order potentials (HOPs), or interactions, among discrete variables. Message passing inference in such models generally takes time exponential in the size of the interaction, but in some cases maximum a posteriori (MAP) inference can be carried out eciently. We build upon such results, introducing two new classes, including composite HOPs that allow us to exibly combine tractable HOPs using simple logical switching rules. We present ecient message update algorithms for the new HOPs, and we improve upon the eciency of message updates for a general class of existing HOPs. Importantly, we present both new and existing HOPs in a common representation; performing inference with any combination of these HOPs requires no change of representations or new derivations.

[1]  Dima Damen,et al.  Computer Vision and Pattern Recognition (CVPR) , 2009 .

[2]  Pushmeet Kohli,et al.  Minimizing sparse higher order energy functions of discrete variables , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Pushmeet Kohli,et al.  P3 & Beyond: Solving Energies with Higher Order Cliques , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[4]  Dima Damen,et al.  Detecting Carried Objects in Short Video Sequences , 2008, ECCV.

[5]  Vladimir Kolmogorov,et al.  Feature Correspondence Via Graph Matching: Models and Global Optimization , 2008, ECCV.

[6]  Martin J. Wainwright,et al.  MAP estimation via agreement on trees: message-passing and linear programming , 2005, IEEE Transactions on Information Theory.

[7]  Brian Potetz,et al.  Efficient Belief Propagation for Vision Using Linear Constraint Nodes , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[8]  Yuri Boykov,et al.  Globally optimal segmentation of multi-region objects , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[9]  Philip H. S. Torr,et al.  Solving Energies with Higher Order Cliques , 2007 .

[10]  Bert Huang,et al.  Loopy Belief Propagation for Bipartite Maximum Weight b-Matching , 2007, AISTATS.

[11]  Brendan J. Frey,et al.  Flexible Priors for Exemplar-based Clustering , 2008, UAI.

[12]  N. Maculan,et al.  Global optimization : from theory to implementation , 2006 .

[13]  Tommi S. Jaakkola,et al.  Fixing Max-Product: Convergent Message Passing Algorithms for MAP LP-Relaxations , 2007, NIPS.

[14]  Brendan J. Frey,et al.  A Binary Variable Model for Affinity Propagation , 2009, Neural Computation.

[15]  Pushmeet Kohli,et al.  Robust Higher Order Potentials for Enforcing Label Consistency , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[16]  Dimitri P. Bertsekas,et al.  Nonlinear Programming , 1997 .

[17]  N. T. Hoai-Phuong,et al.  Optimization under Composite Monotonic Constraints and Constrained Optimization over the Efficient Set , 2006 .

[18]  Rahul Gupta,et al.  Efficient inference with cardinality-based clique potentials , 2007, ICML '07.

[19]  Pushmeet Kohli,et al.  Minimizing sparse higher order energy functions of discrete variables , 2009, CVPR.

[20]  Thomas Hofmann,et al.  Using Combinatorial Optimization within Max-Product Belief Propagation , 2007 .

[21]  Nikos Komodakis,et al.  Beyond pairwise energies: Efficient optimization for higher-order MRFs , 2009, CVPR.

[22]  M. Trick,et al.  Voting schemes for which it can be difficult to tell who won the election , 1989 .

[23]  Keith M. Chugg,et al.  Iterative Message Passing Algorithm for Bipartite Maximum Weighted Matching , 2006, 2006 IEEE International Symposium on Information Theory.

[24]  Craig Boutilier,et al.  Context-Specific Independence in Bayesian Networks , 1996, UAI.

[25]  Vladimir Kolmogorov,et al.  Joint optimization of segmentation and appearance models , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[26]  Tomás Werner,et al.  High-arity interactions, polyhedral relaxations, and cutting plane algorithm for soft constraint optimisation (MAP-MRF) , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.