A Comparative Study of Modern Inference Techniques for Discrete Energy Minimization Problem

Szeliski et al. published an influential study in 2006 on energy minimization methods for Markov Random Fields (MRF). This study provided valuable insights in choosing the best op-timization technique for certain classes of problems. While these insights remain generally useful today, the phe-nomenal success of random field models means that the kinds of inference problems that have to be solved changed signif-icantly. Specifically, the models today often include higher order interactions, flexible connectivity structures, large label-spaces of different cardinalities, or learned energy tables. To reflect these changes, we provide a modernized and enlarged study. We present an empirical comparison of 32 state-of-the-art optimization techniques on a corpus of 2,453 energy minimization instances from diverse applications in computer vision. To ensure reproducibility, we evaluate all methods in the OpenGM 2 framework and report extensive results regard-ing runtime and solution quality. Key insights from our study agree with the results of Szeliski et al. for the types of models they studied. However, on new and challenging types of mod-els our findings disagree and suggest that polyhedral meth-ods and integer programming solvers are competitive in terms of runtime and solution quality over a large range of model types