MPLP++: Fast, Parallel Dual Block-Coordinate Ascent for Dense Graphical Models
This work addresses the need for faster and more efficient MAP solvers in computer vision and bio-imaging applications, representing an incremental improvement over prior methods.
The paper tackled the problem of solving dense graphical models for MAP inference by introducing MPLP++, a new solver based on dual block-coordinate ascent, which significantly outperforms existing solvers like TRWS by a large margin and offers high parallelism for GPU and multi-thread CPU implementations.
Dense, discrete Graphical Models with pairwise potentials are a powerful class of models which are employed in state-of-the-art computer vision and bio-imaging applications. This work introduces a new MAP-solver, based on the popular Dual Block-Coordinate Ascent principle. Surprisingly, by making a small change to the low-performing solver, the Max Product Linear Programming (MPLP) algorithm, we derive the new solver MPLP++ that significantly outperforms all existing solvers by a large margin, including the state-of-the-art solver Tree-Reweighted Sequential (TRWS) message-passing algorithm. Additionally, our solver is highly parallel, in contrast to TRWS, which gives a further boost in performance with the proposed GPU and multi-thread CPU implementations. We verify the superiority of our algorithm on dense problems from publicly available benchmarks, as well, as a new benchmark for 6D Object Pose estimation. We also provide an ablation study with respect to graph density.