Housecraft: Building Houses from Rental Ads and Street Views

Hang Chu; Shenlong Wang; Raquel Urtasun; Sanja Fidler

doi:10.1007/978-3-319-46466-4_30

Housecraft: Building Houses from Rental Ads and Street Views

Hang Chu, Shenlong Wang, Raquel Urtasun, Sanja Fidler

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

Abstract

In this paper, we utilize rental ads to create realistic textured 3D models of building exteriors. In particular, we exploit the address of the property and its floorplan, which are typically available in the ad. The address allows us to extract Google StreetView images around the building, while the building’s floorplan allows for an efficient parametrization of the building in 3D via a small set of random variables. We propose an energy minimization framework which jointly reasons about the height of each floor, the vertical positions of windows and doors, as well as the precise location of the building in the world’s map, by exploiting several geometric and semantic cues from the StreetView imagery. To demonstrate the effectiveness of our approach, we collected a new dataset with 174 houses by crawling a popular rental website. Our experiments show that our approach is able to precisely estimate the geometry and location of the property, and can create realistic 3D building models.

Original language	English (US)
Title of host publication	Computer Vision - 14th European Conference, ECCV 2016, Proceedings
Editors	Bastian Leibe, Jiri Matas, Nicu Sebe, Max Welling
Publisher	Springer
Pages	500-516
Number of pages	17
ISBN (Print)	9783319464657
DOIs	https://doi.org/10.1007/978-3-319-46466-4_30
State	Published - 2016
Externally published	Yes
Event	14th European Conference on Computer Vision, ECCV 2016 - Amsterdam, Netherlands Duration: Oct 11 2016 → Oct 14 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9910 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	14th European Conference on Computer Vision, ECCV 2016
Country/Territory	Netherlands
City	Amsterdam
Period	10/11/16 → 10/14/16

Keywords

3D reconstruction
3D scene understanding
Localization

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Online availability

10.1007/978-3-319-46466-4_30

Library availability

Discover UIUC Full Text

Cite this

Chu, H., Wang, S., Urtasun, R., & Fidler, S. (2016). Housecraft: Building Houses from Rental Ads and Street Views. In B. Leibe, J. Matas, N. Sebe, & M. Welling (Eds.), Computer Vision - 14th European Conference, ECCV 2016, Proceedings (pp. 500-516). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9910 LNCS). Springer. https://doi.org/10.1007/978-3-319-46466-4_30

Housecraft: Building Houses from Rental Ads and Street Views. / Chu, Hang; Wang, Shenlong; Urtasun, Raquel et al.
Computer Vision - 14th European Conference, ECCV 2016, Proceedings. ed. / Bastian Leibe; Jiri Matas; Nicu Sebe; Max Welling. Springer, 2016. p. 500-516 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9910 LNCS).

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

Chu, H, Wang, S, Urtasun, R & Fidler, S 2016, Housecraft: Building Houses from Rental Ads and Street Views. in B Leibe, J Matas, N Sebe & M Welling (eds), Computer Vision - 14th European Conference, ECCV 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9910 LNCS, Springer, pp. 500-516, 14th European Conference on Computer Vision, ECCV 2016, Amsterdam, Netherlands, 10/11/16. https://doi.org/10.1007/978-3-319-46466-4_30

Chu H, Wang S, Urtasun R, Fidler S. Housecraft: Building Houses from Rental Ads and Street Views. In Leibe B, Matas J, Sebe N, Welling M, editors, Computer Vision - 14th European Conference, ECCV 2016, Proceedings. Springer. 2016. p. 500-516. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-46466-4_30

Chu, Hang ; Wang, Shenlong ; Urtasun, Raquel et al. / Housecraft : Building Houses from Rental Ads and Street Views. Computer Vision - 14th European Conference, ECCV 2016, Proceedings. editor / Bastian Leibe ; Jiri Matas ; Nicu Sebe ; Max Welling. Springer, 2016. pp. 500-516 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{1863050023dc4d78b4c78e8fbe757a7c,

title = "Housecraft: Building Houses from Rental Ads and Street Views",

abstract = "In this paper, we utilize rental ads to create realistic textured 3D models of building exteriors. In particular, we exploit the address of the property and its floorplan, which are typically available in the ad. The address allows us to extract Google StreetView images around the building, while the building{\textquoteright}s floorplan allows for an efficient parametrization of the building in 3D via a small set of random variables. We propose an energy minimization framework which jointly reasons about the height of each floor, the vertical positions of windows and doors, as well as the precise location of the building in the world{\textquoteright}s map, by exploiting several geometric and semantic cues from the StreetView imagery. To demonstrate the effectiveness of our approach, we collected a new dataset with 174 houses by crawling a popular rental website. Our experiments show that our approach is able to precisely estimate the geometry and location of the property, and can create realistic 3D building models.",

keywords = "3D reconstruction, 3D scene understanding, Localization",

author = "Hang Chu and Shenlong Wang and Raquel Urtasun and Sanja Fidler",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; 14th European Conference on Computer Vision, ECCV 2016 ; Conference date: 11-10-2016 Through 14-10-2016",

year = "2016",

doi = "10.1007/978-3-319-46466-4_30",

language = "English (US)",

isbn = "9783319464657",

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

publisher = "Springer",

pages = "500--516",

editor = "Bastian Leibe and Jiri Matas and Nicu Sebe and Max Welling",

booktitle = "Computer Vision - 14th European Conference, ECCV 2016, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Housecraft

T2 - 14th European Conference on Computer Vision, ECCV 2016

AU - Chu, Hang

AU - Wang, Shenlong

AU - Urtasun, Raquel

AU - Fidler, Sanja

N1 - Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016

Y1 - 2016

N2 - In this paper, we utilize rental ads to create realistic textured 3D models of building exteriors. In particular, we exploit the address of the property and its floorplan, which are typically available in the ad. The address allows us to extract Google StreetView images around the building, while the building’s floorplan allows for an efficient parametrization of the building in 3D via a small set of random variables. We propose an energy minimization framework which jointly reasons about the height of each floor, the vertical positions of windows and doors, as well as the precise location of the building in the world’s map, by exploiting several geometric and semantic cues from the StreetView imagery. To demonstrate the effectiveness of our approach, we collected a new dataset with 174 houses by crawling a popular rental website. Our experiments show that our approach is able to precisely estimate the geometry and location of the property, and can create realistic 3D building models.

AB - In this paper, we utilize rental ads to create realistic textured 3D models of building exteriors. In particular, we exploit the address of the property and its floorplan, which are typically available in the ad. The address allows us to extract Google StreetView images around the building, while the building’s floorplan allows for an efficient parametrization of the building in 3D via a small set of random variables. We propose an energy minimization framework which jointly reasons about the height of each floor, the vertical positions of windows and doors, as well as the precise location of the building in the world’s map, by exploiting several geometric and semantic cues from the StreetView imagery. To demonstrate the effectiveness of our approach, we collected a new dataset with 174 houses by crawling a popular rental website. Our experiments show that our approach is able to precisely estimate the geometry and location of the property, and can create realistic 3D building models.

KW - 3D reconstruction

KW - 3D scene understanding

KW - Localization

UR - http://www.scopus.com/inward/record.url?scp=84990047389&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84990047389&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-46466-4_30

DO - 10.1007/978-3-319-46466-4_30

M3 - Conference contribution

AN - SCOPUS:84990047389

SN - 9783319464657

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

SP - 500

EP - 516

BT - Computer Vision - 14th European Conference, ECCV 2016, Proceedings

A2 - Leibe, Bastian

A2 - Matas, Jiri

A2 - Sebe, Nicu

A2 - Welling, Max

PB - Springer

Y2 - 11 October 2016 through 14 October 2016

ER -

Housecraft: Building Houses from Rental Ads and Street Views

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this