statefilters

A MATLAB implementation of Learned State Filters.

What is it?

This package is a MATLAB implementation of Learned State Filters for fast discrete pairwise energy minimization as presented
in the following paper:

@InProceedings{Guillaumin2013cvpr,
title = {{Fast Energy Minimization using Learned State Filters}},
author = {Guillaumin, Matthieu and Van Gool, Luc and Ferrari, Vittorio},
booktitle = {{IEEE Conference on Computer Vision \& Pattern Recognition (CVPR)}},
year = {2013},
month = jun,
pdf = {http://biwinas03.ee.ethz.ch/mguillau/publications/Guillaumin2013cvpr.pdf}
}

Please cite this work if you use this package for publication.

For convenient comparison to other energy minimization work, we have also included code for TRW-S, alpha-expansion, alpha-expand beta-shrink and loopy belief propagation.

How to install?

1. First, download and install Matlabtools: bitbucket:matlabtools.

2. To activate it in Matlab, run:

   run('/path/to/matlabtools/setup.m');
   

3. Optionally, download pre-computed data and models from here.

How to use?

After you have followed the installation and configuration steps above, you
simply need to run setup.m in MATLAB:

run('/path/to/statefilters/setup.m');

Then you can run the simple demo code:

demo();

The main entry point for this toolbox is the gm_solve function. Usage of this function is described in its help section and demo.m is an example use of TRW-S and state filters on pre-computed data and models.