<brussell@csail.mit.edu>
<torralba@csail.mit.edu>


LabelMe3D Matlab toolbox


CITATION:

B. C. Russell and A. Torralba
Building a Database of 3D Scenes from User Annotations
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2009


DESCRIPTION:

Here you will find the LabelMe3D Matlab toolbox.  Given a LabelMe
annotation file, the toolbox contains source code to infer world
3D coordinates for labeled objects in the scene and recover the
camera parameters for the image.  Moreover, there are functions for
visualizing the inferred 3D information.


QUICK START:

demo.m - This file has a demonstration of the main features of the
toolbox.  It is best to read and step through the demonstration
line-by-line.

Documentation.pdf - This file documents the scene geometry and format
of the annotation XML files.


(c) 2013, MIT Computer Science and Artificial Intelligence Laboratory