Designing deep networks for surface normal estimation

Xiaolong Wang; David F. Fouhey; Abhinav Gupta

doi:10.1109/CVPR.2015.7298652

Designing deep networks for surface normal estimation

Xiaolong Wang, David F. Fouhey, Abhinav Gupta

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In the past few years, convolutional neural nets (CNN) have shown incredible promise for learning visual representations. In this paper, we use CNNs for the task of predicting surface normals from a single image. But what is the right architecture? We propose to build upon the decades of hard work in 3D scene understanding to design a new CNN architecture for the task of surface normal estimation. We show that incorporating several constraints (man-made, Manhattan world) and meaningful intermediate representations (room layout, edge labels) in the architecture leads to state of the art performance on surface normal estimation. We also show that our network is quite robust and show state of the art results on other datasets as well without any fine-tuning.

Original language	English (US)
Title of host publication	IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015
Publisher	IEEE Computer Society
Pages	539-547
Number of pages	9
ISBN (Electronic)	9781467369640
DOIs	https://doi.org/10.1109/CVPR.2015.7298652
State	Published - Oct 14 2015
Event	IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015 - Boston, United States Duration: Jun 7 2015 → Jun 12 2015

Publication series

Name	Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Volume	07-12-June-2015
ISSN (Print)	1063-6919

Other

Other	IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015
Country/Territory	United States
City	Boston
Period	6/7/15 → 6/12/15

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition

Access to Document

10.1109/CVPR.2015.7298652

Cite this

Wang, X., Fouhey, D. F., & Gupta, A. (2015). Designing deep networks for surface normal estimation. In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015 (pp. 539-547). Article 7298652 (Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition; Vol. 07-12-June-2015). IEEE Computer Society. https://doi.org/10.1109/CVPR.2015.7298652

Designing deep networks for surface normal estimation. / Wang, Xiaolong; Fouhey, David F.; Gupta, Abhinav.
IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015. IEEE Computer Society, 2015. p. 539-547 7298652 (Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition; Vol. 07-12-June-2015).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Wang, X, Fouhey, DF & Gupta, A 2015, Designing deep networks for surface normal estimation. in IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015., 7298652, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 07-12-June-2015, IEEE Computer Society, pp. 539-547, IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015, Boston, United States, 6/7/15. https://doi.org/10.1109/CVPR.2015.7298652

@inproceedings{e268b39052f842da9bb6510521ee3225,

title = "Designing deep networks for surface normal estimation",

abstract = "In the past few years, convolutional neural nets (CNN) have shown incredible promise for learning visual representations. In this paper, we use CNNs for the task of predicting surface normals from a single image. But what is the right architecture? We propose to build upon the decades of hard work in 3D scene understanding to design a new CNN architecture for the task of surface normal estimation. We show that incorporating several constraints (man-made, Manhattan world) and meaningful intermediate representations (room layout, edge labels) in the architecture leads to state of the art performance on surface normal estimation. We also show that our network is quite robust and show state of the art results on other datasets as well without any fine-tuning.",

author = "Xiaolong Wang and Fouhey, {David F.} and Abhinav Gupta",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015 ; Conference date: 07-06-2015 Through 12-06-2015",

year = "2015",

month = oct,

day = "14",

doi = "10.1109/CVPR.2015.7298652",

language = "English (US)",

series = "Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition",

publisher = "IEEE Computer Society",

pages = "539--547",

booktitle = "IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015",

}

TY - GEN

T1 - Designing deep networks for surface normal estimation

AU - Wang, Xiaolong

AU - Fouhey, David F.

AU - Gupta, Abhinav

PY - 2015/10/14

Y1 - 2015/10/14

N2 - In the past few years, convolutional neural nets (CNN) have shown incredible promise for learning visual representations. In this paper, we use CNNs for the task of predicting surface normals from a single image. But what is the right architecture? We propose to build upon the decades of hard work in 3D scene understanding to design a new CNN architecture for the task of surface normal estimation. We show that incorporating several constraints (man-made, Manhattan world) and meaningful intermediate representations (room layout, edge labels) in the architecture leads to state of the art performance on surface normal estimation. We also show that our network is quite robust and show state of the art results on other datasets as well without any fine-tuning.

AB - In the past few years, convolutional neural nets (CNN) have shown incredible promise for learning visual representations. In this paper, we use CNNs for the task of predicting surface normals from a single image. But what is the right architecture? We propose to build upon the decades of hard work in 3D scene understanding to design a new CNN architecture for the task of surface normal estimation. We show that incorporating several constraints (man-made, Manhattan world) and meaningful intermediate representations (room layout, edge labels) in the architecture leads to state of the art performance on surface normal estimation. We also show that our network is quite robust and show state of the art results on other datasets as well without any fine-tuning.

UR - http://www.scopus.com/inward/record.url?scp=84959234840&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959234840&partnerID=8YFLogxK

U2 - 10.1109/CVPR.2015.7298652

DO - 10.1109/CVPR.2015.7298652

M3 - Conference contribution

AN - SCOPUS:84959234840

T3 - Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

SP - 539

EP - 547

BT - IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015

PB - IEEE Computer Society

T2 - IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015

Y2 - 7 June 2015 through 12 June 2015

ER -

Designing deep networks for surface normal estimation

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this