Welcome, this is the C++ code corresponding to "Parsing Geometry Using Structure Aware Shape Templates".
This is a tough pipeline and therefore, multiple parts of the code are maintained here, therefore making it one large clutter.
Under these conditions, this code is expected to run successfully: MS Visual Studio Express 2013, glew-1.12.0, freeglut-3.0, flann-1.7.1
First start by running build.bat after editing main.pro. Make edits to main.pro to reflect onto the corresponding paths, and this can be made to work.
The required templates are present in Data/Grammars.
To run the optimization, one would need to capture the point cloud from the mesh by pressing "X" on the keyboard. Multiple view point capture can be obtained
by pressing "X" and moving the mouse around with a long press. Pressing "O" and running it on mode "0" will start the optimization with a decent initialization.
Any change to code can be effected on Scene.cpp (for the core/changing or manipulating shapes), Grammar.cpp (for the templates, changing/manipulating them), cmaes.h (for the CMA-ES optimization) and GLWidget.cpp/GUI.cpp (for the key-press operations).
Any use of this code in any publication, patent or for any other research purposes, requires citing of the following paper:
Ganapathi-Subramanian, V., Diamanti, O., Pirk, S., Tang, C., Niessner, M. and Guibas, L., 2018, September. Parsing geometry using structure-aware shape templates. In 2018 International Conference on 3D Vision (3DV) (pp. 672-681). IEEE.
The BibTeX for the same is as follows:
@inproceedings{ganapathi2018parsing,
title={Parsing geometry using structure-aware shape templates},
author={Ganapathi-Subramanian, Vignesh and Diamanti, Olga and Pirk, Soeren and Tang, Chengcheng and Niessner, Matthias and Guibas, Leonidas},
booktitle={2018 International Conference on 3D Vision (3DV)},
pages={672--681},
year={2018},
organization={IEEE}
}