See:
license.m, 
LearningCode/README

Version 0.1

Running OneShot3dEfficent the First Time
----------------------------------------

-Add the line "addpath(genpath('../ec2/bin/mex'));" to LearningCode/InitialPath.m, if it is not already there
-start matlab
-cd to LearningCode
-Run InitialPath
-Run "OneShot3dEfficient("input_image_name.jpg", "output_dir")




A few points to note:
1.  The code will not run out of the box.  Therefore, unless
    you have read the ICCV-3dRR paper, it would be hard to 
    make it run.
2.  This code may differe from the one used for experiments 
    in ICCV-3dRR.
3.  (For our recent submission for ICCV-3dRR + ICCV-vrml + NIPS 
    workshop on grammar of vision work, to IEEE-PAMI, please email us.)
4.  For understanding the optimization used for learning and inference
    in the paper (minimization of L1 norms), please first see:
    http://www.stanford.edu/class/ee364a/lectures.html
    (lecture 4 on convex optimization problems)
    before emailing us.



Useful links:
http://make3d.stanford.edu/publications
http://make3d.stanford.edu/publications/code
http://make3d.stanford.edu/publications/faq
http://make3d.stanford.edu/publications/contact