A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery

Nathan Jacobs; Adam Kraft; Muhammad Usman Rafique; Ranti Dev Sharma

doi:10.1145/3274895.3274934

A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery

Nathan Jacobs, Adam Kraft, Muhammad Usman Rafique, Ranti Dev Sharma

Computer Science

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

9 Scopus citations

Abstract

We propose a neural network component, the regional aggregation layer, that makes it possible to train a pixel-level density estimator using only coarse-grained density aggregates, which reflect the number of objects in an image region. Our approach is simple to use and does not require domain-specific assumptions about the nature of the density function. We evaluate our approach on several synthetic datasets. In addition, we use this approach to learn to estimate high-resolution population and housing density from satellite imagery. In all cases, we find that our approach results in better density estimates than a commonly used baseline. We also show how our housing density estimator can be used to classify buildings as residential or non-residential.

Original language	English
Title of host publication	26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018
Editors	Li Xiong, Roberto Tamassia, Kashani Farnoush Banaei, Ralf Hartmut Guting, Erik Hoel
Pages	33-42
Number of pages	10
ISBN (Electronic)	9781450358897
DOIs	https://doi.org/10.1145/3274895.3274934
State	Published - Nov 6 2018
Event	26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018 - Seattle, United States Duration: Nov 6 2018 → Nov 9 2018

Publication series

Name	GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems

Conference

Conference	26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018
Country/Territory	United States
City	Seattle
Period	11/6/18 → 11/9/18

Bibliographical note

Publisher Copyright:
© 2018 Association for Computing Machinery.

Keywords

Dasymetric mapping
Population density
Remote sensing

ASJC Scopus subject areas

Earth-Surface Processes
Computer Science Applications
Modeling and Simulation
Computer Graphics and Computer-Aided Design
Information Systems

Access to Document

10.1145/3274895.3274934

Cite this

Jacobs, N., Kraft, A., Rafique, M. U., & Sharma, R. D. (2018). A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery. In L. Xiong, R. Tamassia, K. F. Banaei, R. H. Guting, & E. Hoel (Eds.), 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018 (pp. 33-42). (GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems). https://doi.org/10.1145/3274895.3274934

Jacobs, Nathan ; Kraft, Adam ; Rafique, Muhammad Usman et al. / A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery. 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018. editor / Li Xiong ; Roberto Tamassia ; Kashani Farnoush Banaei ; Ralf Hartmut Guting ; Erik Hoel. 2018. pp. 33-42 (GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems).

@inproceedings{af98e960a540489981ae48672aef3ad9,

title = "A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery",

abstract = "We propose a neural network component, the regional aggregation layer, that makes it possible to train a pixel-level density estimator using only coarse-grained density aggregates, which reflect the number of objects in an image region. Our approach is simple to use and does not require domain-specific assumptions about the nature of the density function. We evaluate our approach on several synthetic datasets. In addition, we use this approach to learn to estimate high-resolution population and housing density from satellite imagery. In all cases, we find that our approach results in better density estimates than a commonly used baseline. We also show how our housing density estimator can be used to classify buildings as residential or non-residential.",

keywords = "Dasymetric mapping, Population density, Remote sensing",

author = "Nathan Jacobs and Adam Kraft and Rafique, {Muhammad Usman} and Sharma, {Ranti Dev}",

note = "Publisher Copyright: {\textcopyright} 2018 Association for Computing Machinery.; 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018 ; Conference date: 06-11-2018 Through 09-11-2018",

year = "2018",

month = nov,

day = "6",

doi = "10.1145/3274895.3274934",

language = "English",

series = "GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems",

pages = "33--42",

editor = "Li Xiong and Roberto Tamassia and Banaei, {Kashani Farnoush} and Guting, {Ralf Hartmut} and Erik Hoel",

booktitle = "26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018",

}

Jacobs, N, Kraft, A, Rafique, MU & Sharma, RD 2018, A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery. in L Xiong, R Tamassia, KF Banaei, RH Guting & E Hoel (eds), 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018. GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems, pp. 33-42, 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018, Seattle, United States, 11/6/18. https://doi.org/10.1145/3274895.3274934

A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery. / Jacobs, Nathan; Kraft, Adam; Rafique, Muhammad Usman et al.
26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018. ed. / Li Xiong; Roberto Tamassia; Kashani Farnoush Banaei; Ralf Hartmut Guting; Erik Hoel. 2018. p. 33-42 (GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems).

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

TY - GEN

T1 - A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery

AU - Jacobs, Nathan

AU - Kraft, Adam

AU - Rafique, Muhammad Usman

AU - Sharma, Ranti Dev

PY - 2018/11/6

Y1 - 2018/11/6

N2 - We propose a neural network component, the regional aggregation layer, that makes it possible to train a pixel-level density estimator using only coarse-grained density aggregates, which reflect the number of objects in an image region. Our approach is simple to use and does not require domain-specific assumptions about the nature of the density function. We evaluate our approach on several synthetic datasets. In addition, we use this approach to learn to estimate high-resolution population and housing density from satellite imagery. In all cases, we find that our approach results in better density estimates than a commonly used baseline. We also show how our housing density estimator can be used to classify buildings as residential or non-residential.

AB - We propose a neural network component, the regional aggregation layer, that makes it possible to train a pixel-level density estimator using only coarse-grained density aggregates, which reflect the number of objects in an image region. Our approach is simple to use and does not require domain-specific assumptions about the nature of the density function. We evaluate our approach on several synthetic datasets. In addition, we use this approach to learn to estimate high-resolution population and housing density from satellite imagery. In all cases, we find that our approach results in better density estimates than a commonly used baseline. We also show how our housing density estimator can be used to classify buildings as residential or non-residential.

KW - Dasymetric mapping

KW - Population density

KW - Remote sensing

UR - http://www.scopus.com/inward/record.url?scp=85058635628&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85058635628&partnerID=8YFLogxK

U2 - 10.1145/3274895.3274934

DO - 10.1145/3274895.3274934

M3 - Conference contribution

AN - SCOPUS:85058635628

T3 - GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems

SP - 33

EP - 42

BT - 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018

A2 - Xiong, Li

A2 - Tamassia, Roberto

A2 - Banaei, Kashani Farnoush

A2 - Guting, Ralf Hartmut

A2 - Hoel, Erik

T2 - 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018

Y2 - 6 November 2018 through 9 November 2018

ER -

Jacobs N, Kraft A, Rafique MU, Sharma RD. A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery. In Xiong L, Tamassia R, Banaei KF, Guting RH, Hoel E, editors, 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2018. 2018. p. 33-42. (GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems). doi: 10.1145/3274895.3274934

A weakly supervised approach for estimating spatial density functions from high-resolution satellite imagery

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this