This repo some of the data pertaining to the TOG'14 paper:

Kevin Karsch, Kalyan Sunkavalli, Sunil Hadap, Nathan Carr, Hailin Jin, Rafael Fonte, Michael Sittig, David A Forsyth. "Automatic scene inference for 3D object compositing." TOG 2014

There are two folders:

IBL-Scenes/ - four 3D scenes with ground truth lighting/reflectance and so on
StudySynthetic/ - all data used for evaluations in the paper