A new look at reweighted message passing
This work provides a simpler framework for message passing in graphical models, potentially benefiting researchers and practitioners in machine learning and computer vision, though it appears incremental.
The authors tackled the problem of MAP estimation in graphical models by proposing Sequential Reweighted Message Passing (SRMP), which simplifies the derivation of existing methods like TRW-S and enables generalizations to higher-order models, achieving promising results on real-world problems.
We propose a new family of message passing techniques for MAP estimation in graphical models which we call {\em Sequential Reweighted Message Passing} (SRMP). Special cases include well-known techniques such as {\em Min-Sum Diffusion} (MSD) and a faster {\em Sequential Tree-Reweighted Message Passing} (TRW-S). Importantly, our derivation is simpler than the original derivation of TRW-S, and does not involve a decomposition into trees. This allows easy generalizations. We present such a generalization for the case of higher-order graphical models, and test it on several real-world problems with promising results.