Single Image 3D Without a Single 3D Image

David F. Fouhey, Wajahat Hussain, Abhinav Gupta, Martial Hebert; The IEEE International Conference on Computer Vision (ICCV), 2015, pp. 1053-1061

Abstract


Do we really need 3D labels in order to learn how to predict 3D? In this paper, we show that one can learn a mapping from appearance to 3D properties without ever seeing a single explicit 3D label. Rather than use explicit supervision, we use the regularity of indoor scenes to learn the mapping in a completely unsupervised manner. We demonstrate this on both a standard 3D scene understanding dataset as well as Internet images for which 3D is unavailable, precluding supervised learning. Despite never seeing a 3D label, our method produces competitive results.

Related Material


[pdf]
[bibtex]
@InProceedings{Fouhey_2015_ICCV,
author = {Fouhey, David F. and Hussain, Wajahat and Gupta, Abhinav and Hebert, Martial},
title = {Single Image 3D Without a Single 3D Image},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {December},
year = {2015}
}