General 3D Room Layout from a Single View by Render-and-Compare

Sinisa Stekovic; Shreyas Hampali Shivakumar; Mahdi Rad; Sayan Deb Sarkar; Friedrich Fraundorfer; Vincent Lepetit

doi:10.1007/978-3-030-58517-4_12

General 3D Room Layout from a Single View by Render-and-Compare

Sinisa Stekovic^*, Shreyas Hampali Shivakumar, Mahdi Rad, Sayan Deb Sarkar, Friedrich Fraundorfer, Vincent Lepetit

^*Corresponding author for this work

Institute of Computer Graphics and Vision (7100)

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

We present a novel method to reconstruct the 3D layout of a room—walls, floors, ceilings—from a single perspective view in challenging conditions, by contrast with previous single-view methods restricted to cuboid-shaped layouts. This input view can consist of a color image only, but considering a depth map results in a more accurate reconstruction. Our approach is formalized as solving a constrained discrete optimization problem to find the set of 3D polygons that constitute the layout. In order to deal with occlusions between components of the layout, which is a problem ignored by previous works, we introduce an analysis-by-synthesis method to iteratively refine the 3D layout estimate. As no dataset was available to evaluate our method quantitatively, we created one together with several appropriate metrics. Our dataset consists of 293 images from ScanNet, which we annotated with precise 3D layouts. It offers three times more samples than the popular NYUv2 303 benchmark, and a much larger variety of layouts.

Original language	English
Title of host publication	Computer Vision – ECCV 2020 - 16th European Conference, Proceedings
Editors	Andrea Vedaldi, Horst Bischof, Thomas Brox, Jan-Michael Frahm
Publisher	Springer, Cham
Pages	187-203
Number of pages	17
ISBN (Print)	9783030585167
DOIs	https://doi.org/10.1007/978-3-030-58517-4_12
Publication status	Published - 23 Aug 2020
Event	16th European Conference on Computer Vision: ECCV 2020 - Virtual, Glasgow, United Kingdom Duration: 23 Aug 2020 → 28 Aug 2020

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12361 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th European Conference on Computer Vision
Abbreviated title	ECCV 2020
Country/Territory	United Kingdom
City	Virtual, Glasgow
Period	23/08/20 → 28/08/20

Keywords

3D geometry
Analysis-by-synthesis
Room layout

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-030-58517-4_12

Cite this

Stekovic, S., Hampali Shivakumar, S., Rad, M., Deb Sarkar, S., Fraundorfer, F., & Lepetit, V. (2020). General 3D Room Layout from a Single View by Render-and-Compare. In A. Vedaldi, H. Bischof, T. Brox, & J-M. Frahm (Eds.), Computer Vision – ECCV 2020 - 16th European Conference, Proceedings (pp. 187-203). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12361 LNCS). Springer, Cham. https://doi.org/10.1007/978-3-030-58517-4_12

General 3D Room Layout from a Single View by Render-and-Compare. / Stekovic, Sinisa; Hampali Shivakumar, Shreyas; Rad, Mahdi et al.
Computer Vision – ECCV 2020 - 16th European Conference, Proceedings. ed. / Andrea Vedaldi; Horst Bischof; Thomas Brox; Jan-Michael Frahm. Springer, Cham, 2020. p. 187-203 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12361 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Stekovic, S, Hampali Shivakumar, S, Rad, M, Deb Sarkar, S, Fraundorfer, F & Lepetit, V 2020, General 3D Room Layout from a Single View by Render-and-Compare. in A Vedaldi, H Bischof, T Brox & J-M Frahm (eds), Computer Vision – ECCV 2020 - 16th European Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12361 LNCS, Springer, Cham, pp. 187-203, 16th European Conference on Computer Vision, Virtual, Glasgow, United Kingdom, 23/08/20. https://doi.org/10.1007/978-3-030-58517-4_12

Stekovic S, Hampali Shivakumar S, Rad M, Deb Sarkar S, Fraundorfer F, Lepetit V. General 3D Room Layout from a Single View by Render-and-Compare. In Vedaldi A, Bischof H, Brox T, Frahm J-M, editors, Computer Vision – ECCV 2020 - 16th European Conference, Proceedings. Springer, Cham. 2020. p. 187-203. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-58517-4_12

Stekovic, Sinisa ; Hampali Shivakumar, Shreyas ; Rad, Mahdi et al. / General 3D Room Layout from a Single View by Render-and-Compare. Computer Vision – ECCV 2020 - 16th European Conference, Proceedings. editor / Andrea Vedaldi ; Horst Bischof ; Thomas Brox ; Jan-Michael Frahm. Springer, Cham, 2020. pp. 187-203 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{f35fc132e13e4a7ea84d209991f99611,

title = "General 3D Room Layout from a Single View by Render-and-Compare",

abstract = "We present a novel method to reconstruct the 3D layout of a room—walls, floors, ceilings—from a single perspective view in challenging conditions, by contrast with previous single-view methods restricted to cuboid-shaped layouts. This input view can consist of a color image only, but considering a depth map results in a more accurate reconstruction. Our approach is formalized as solving a constrained discrete optimization problem to find the set of 3D polygons that constitute the layout. In order to deal with occlusions between components of the layout, which is a problem ignored by previous works, we introduce an analysis-by-synthesis method to iteratively refine the 3D layout estimate. As no dataset was available to evaluate our method quantitatively, we created one together with several appropriate metrics. Our dataset consists of 293 images from ScanNet, which we annotated with precise 3D layouts. It offers three times more samples than the popular NYUv2 303 benchmark, and a much larger variety of layouts.",

keywords = "3D geometry, Analysis-by-synthesis, Room layout",

author = "Sinisa Stekovic and {Hampali Shivakumar}, Shreyas and Mahdi Rad and {Deb Sarkar}, Sayan and Friedrich Fraundorfer and Vincent Lepetit",

year = "2020",

month = aug,

day = "23",

doi = "10.1007/978-3-030-58517-4_12",

language = "English",

isbn = "9783030585167",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer, Cham",

pages = "187--203",

editor = "Andrea Vedaldi and Horst Bischof and Thomas Brox and Jan-Michael Frahm",

booktitle = "Computer Vision – ECCV 2020 - 16th European Conference, Proceedings",

note = "16th European Conference on Computer Vision : ECCV 2020, ECCV 2020 ; Conference date: 23-08-2020 Through 28-08-2020",

}

TY - GEN

T1 - General 3D Room Layout from a Single View by Render-and-Compare

AU - Stekovic, Sinisa

AU - Hampali Shivakumar, Shreyas

AU - Rad, Mahdi

AU - Deb Sarkar, Sayan

AU - Fraundorfer, Friedrich

AU - Lepetit, Vincent

PY - 2020/8/23

Y1 - 2020/8/23

N2 - We present a novel method to reconstruct the 3D layout of a room—walls, floors, ceilings—from a single perspective view in challenging conditions, by contrast with previous single-view methods restricted to cuboid-shaped layouts. This input view can consist of a color image only, but considering a depth map results in a more accurate reconstruction. Our approach is formalized as solving a constrained discrete optimization problem to find the set of 3D polygons that constitute the layout. In order to deal with occlusions between components of the layout, which is a problem ignored by previous works, we introduce an analysis-by-synthesis method to iteratively refine the 3D layout estimate. As no dataset was available to evaluate our method quantitatively, we created one together with several appropriate metrics. Our dataset consists of 293 images from ScanNet, which we annotated with precise 3D layouts. It offers three times more samples than the popular NYUv2 303 benchmark, and a much larger variety of layouts.

AB - We present a novel method to reconstruct the 3D layout of a room—walls, floors, ceilings—from a single perspective view in challenging conditions, by contrast with previous single-view methods restricted to cuboid-shaped layouts. This input view can consist of a color image only, but considering a depth map results in a more accurate reconstruction. Our approach is formalized as solving a constrained discrete optimization problem to find the set of 3D polygons that constitute the layout. In order to deal with occlusions between components of the layout, which is a problem ignored by previous works, we introduce an analysis-by-synthesis method to iteratively refine the 3D layout estimate. As no dataset was available to evaluate our method quantitatively, we created one together with several appropriate metrics. Our dataset consists of 293 images from ScanNet, which we annotated with precise 3D layouts. It offers three times more samples than the popular NYUv2 303 benchmark, and a much larger variety of layouts.

KW - 3D geometry

KW - Analysis-by-synthesis

KW - Room layout

UR - http://www.scopus.com/inward/record.url?scp=85092935640&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-58517-4_12

DO - 10.1007/978-3-030-58517-4_12

M3 - Conference paper

SN - 9783030585167

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 187

EP - 203

BT - Computer Vision – ECCV 2020 - 16th European Conference, Proceedings

A2 - Vedaldi, Andrea

A2 - Bischof, Horst

A2 - Brox, Thomas

A2 - Frahm, Jan-Michael

PB - Springer, Cham

T2 - 16th European Conference on Computer Vision

Y2 - 23 August 2020 through 28 August 2020

ER -

General 3D Room Layout from a Single View by Render-and-Compare

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this