/upgmpp

Undirected Probailistic Graphical Modes for C++

Primary LanguageC++GNU General Public License v3.0GPL-3.0

UPGM++

Undirected Probabilistic Graphical Models in C++

Copyright (C) 2014-2017 Jose Raul Ruiz Sarmiento

University of Malaga (jotaraul@uma.es)

Personal webpage: http://mapir.isa.uma.es/jotaraul

Project webpage: http://mapir.isa.uma.es/work/upgmpp-library

If you use UPGM++, please cite us by:

@INPROCEEDINGS{Ruiz-Sarmiento-REACTS-2015, author = {Ruiz-Sarmiento, J. R. and Galindo, Cipriano and Gonz{\'{a}}lez-Jim{\'{e}}nez, Javier}, title = {UPGMpp: a Software Library for Contextual Object Recognition}, booktitle = {3rd. Workshop on Recognition and Action for Scene Understanding (REACTS)}, year = {2015} }

General information

This is a standalone C++ library for Undirected Probabilistic Graphical Models. Some examples are provided for a better understanding of the library use.

I want to thank Prof. Dr. Joachim Hertzberg and Mr. Martin Günther for the motivation and support that they gave me in the first steps of this library, which was started during my stay at the University of Osnabrück, Germany.

Prerequisites

Install dependencies:

  • Eigen.

  • Boost.

  • liblbfgs:

      sudo apt-get install liblbfgs-dev

    That package is available starting with Ubuntu 12.10 Raring; if you're using Ubuntu Precise, go to http://packages.ubuntu.com and download the Raring package (or later). You also need the dependency liblbfgs0. From the command line:

      cd /tmp
  wget http://de.archive.ubuntu.com/ubuntu/pool/universe/libl/liblbfgs/liblbfgs0_1.10-5_amd64.deb     # replace amd64 by i386 if compiling on a 32 bit Ubuntu
  wget http://de.archive.ubuntu.com/ubuntu/pool/universe/libl/liblbfgs/liblbfgs-dev_1.10-5_amd64.deb  # replace amd64 by i386 if compiling on a 32 bit Ubuntu
  sudo dpkg -i liblbfgs0_1.10-5_amd64.deb liblbfgs-dev_1.10-5_amd64.deb

Compiling

mkdir build && cd build && cmake .. && make

Installation

sudo make install