Cross-Modal Learning of Housing Quality in Amsterdam

Alex Levering; Diego Marcos; Ilan Havinga; Devis Tuia

doi:10.1145/3486635.3491067

Cross-Modal Learning of Housing Quality in Amsterdam

Alex Levering, Diego Marcos, Ilan Havinga, Devis Tuia

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

Abstract

In our research we test data and models for the recognition of housing quality in the city of Amsterdam from ground-level and aerial imagery. For ground-level images we compare Google StreetView (GSV) to Flickr images. Our results show that GSV predicts the most accurate building quality scores, approximately 30% better than using only aerial images. However, we find that through careful filtering and by using the right pre-trained model, Flickr image features combined with aerial image features are able to halve the performance gap to GSV features from 30% to 15%. Our results indicate that there are viable alternatives to GSV for liveability factor prediction, which is encouraging as GSV images are more difficult to acquire and not always available.

Original language	English
Title of host publication	Proceedings of the 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021
Editors	D. Lunga, L. Yang, S. Gao, B. Martins, Y. Hu, X. Deng, S. Newsam
Publisher	Association for Computing Machinery, Inc
Pages	1-4
ISBN (Electronic)	9781450391207
DOIs	https://doi.org/10.1145/3486635.3491067
Publication status	Published - 2 Nov 2021
Externally published	Yes
Event	4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021 - Beijing, China Duration: 2 Nov 2021 → 2 Nov 2021

Conference

Conference	4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021
Country/Territory	China
City	Beijing
Period	2/11/21 → 2/11/21

Access to Document

10.1145/3486635.3491067

Persistent URL (handle)

Persistent link

Cite this

Levering, A., Marcos, D., Havinga, I., & Tuia, D. (2021). Cross-Modal Learning of Housing Quality in Amsterdam. In D. Lunga, L. Yang, S. Gao, B. Martins, Y. Hu, X. Deng, & S. Newsam (Eds.), Proceedings of the 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021 (pp. 1-4). Association for Computing Machinery, Inc. https://doi.org/10.1145/3486635.3491067

@inproceedings{c61bbdf8f12844539d6381dfaf38908d,

title = "Cross-Modal Learning of Housing Quality in Amsterdam",

abstract = "In our research we test data and models for the recognition of housing quality in the city of Amsterdam from ground-level and aerial imagery. For ground-level images we compare Google StreetView (GSV) to Flickr images. Our results show that GSV predicts the most accurate building quality scores, approximately 30% better than using only aerial images. However, we find that through careful filtering and by using the right pre-trained model, Flickr image features combined with aerial image features are able to halve the performance gap to GSV features from 30% to 15%. Our results indicate that there are viable alternatives to GSV for liveability factor prediction, which is encouraging as GSV images are more difficult to acquire and not always available.",

author = "Alex Levering and Diego Marcos and Ilan Havinga and Devis Tuia",

year = "2021",

month = nov,

day = "2",

doi = "10.1145/3486635.3491067",

language = "English",

pages = "1--4",

editor = "D. Lunga and L. Yang and S. Gao and B. Martins and Y. Hu and X. Deng and S. Newsam",

booktitle = "Proceedings of the 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021",

publisher = "Association for Computing Machinery, Inc",

note = "4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021 ; Conference date: 02-11-2021 Through 02-11-2021",

}

Levering, A, Marcos, D, Havinga, I & Tuia, D 2021, Cross-Modal Learning of Housing Quality in Amsterdam. in D Lunga, L Yang, S Gao, B Martins, Y Hu, X Deng & S Newsam (eds), Proceedings of the 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021. Association for Computing Machinery, Inc, pp. 1-4, 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021, Beijing, China, 2/11/21. https://doi.org/10.1145/3486635.3491067

Cross-Modal Learning of Housing Quality in Amsterdam. / Levering, Alex; Marcos, Diego; Havinga, Ilan et al.
Proceedings of the 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021. ed. / D. Lunga; L. Yang; S. Gao; B. Martins; Y. Hu; X. Deng; S. Newsam. Association for Computing Machinery, Inc, 2021. p. 1-4.

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

TY - GEN

T1 - Cross-Modal Learning of Housing Quality in Amsterdam

AU - Levering, Alex

AU - Marcos, Diego

AU - Havinga, Ilan

AU - Tuia, Devis

PY - 2021/11/2

Y1 - 2021/11/2

N2 - In our research we test data and models for the recognition of housing quality in the city of Amsterdam from ground-level and aerial imagery. For ground-level images we compare Google StreetView (GSV) to Flickr images. Our results show that GSV predicts the most accurate building quality scores, approximately 30% better than using only aerial images. However, we find that through careful filtering and by using the right pre-trained model, Flickr image features combined with aerial image features are able to halve the performance gap to GSV features from 30% to 15%. Our results indicate that there are viable alternatives to GSV for liveability factor prediction, which is encouraging as GSV images are more difficult to acquire and not always available.

AB - In our research we test data and models for the recognition of housing quality in the city of Amsterdam from ground-level and aerial imagery. For ground-level images we compare Google StreetView (GSV) to Flickr images. Our results show that GSV predicts the most accurate building quality scores, approximately 30% better than using only aerial images. However, we find that through careful filtering and by using the right pre-trained model, Flickr image features combined with aerial image features are able to halve the performance gap to GSV features from 30% to 15%. Our results indicate that there are viable alternatives to GSV for liveability factor prediction, which is encouraging as GSV images are more difficult to acquire and not always available.

UR - http://www.scopus.com/inward/record.url?scp=85119498139&partnerID=8YFLogxK

U2 - 10.1145/3486635.3491067

DO - 10.1145/3486635.3491067

M3 - Conference contribution

SP - 1

EP - 4

BT - Proceedings of the 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021

A2 - Lunga, D.

A2 - Yang, L.

A2 - Gao, S.

A2 - Martins, B.

A2 - Hu, Y.

A2 - Deng, X.

A2 - Newsam, S.

PB - Association for Computing Machinery, Inc

T2 - 4th ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery, GeoAI 2021

Y2 - 2 November 2021 through 2 November 2021

ER -

Cross-Modal Learning of Housing Quality in Amsterdam

Abstract

Conference

Access to Document

Persistent URL (handle)

Other files and links

Fingerprint

Cite this