Home > Research > Publications & Outputs > Disaster mapping from satellites

Research

Associated organisational unit

Electronic data

View graph of relations

Disaster mapping from satellites: damage detection with crowdsourced point labels

Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review

Published

Standard

Disaster mapping from satellites: damage detection with crowdsourced point labels. / Kuzin, Danil; Simmons, Brooke; Isupova, Olga et al.
2021. Paper presented at NeurIPS 2021.

Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review

Harvard

Kuzin, D, Simmons, B, Isupova, O & Reece, S 2021, 'Disaster mapping from satellites: damage detection with crowdsourced point labels', Paper presented at NeurIPS 2021, 6/12/21 - 14/12/21.

APA

Kuzin, D., Simmons, B., Isupova, O., & Reece, S. (2021). Disaster mapping from satellites: damage detection with crowdsourced point labels. Paper presented at NeurIPS 2021.

Vancouver

Kuzin D, Simmons B, Isupova O, Reece S. Disaster mapping from satellites: damage detection with crowdsourced point labels. 2021. Paper presented at NeurIPS 2021.

Author

Kuzin, Danil ; Simmons, Brooke ; Isupova, Olga et al. / Disaster mapping from satellites : damage detection with crowdsourced point labels. Paper presented at NeurIPS 2021.10 p.

Bibtex

@conference{54778bbe39724526ac1e2b7224df08c2,
title = "Disaster mapping from satellites: damage detection with crowdsourced point labels",
abstract = "High-resolution satellite imagery available immediately after disaster events is crucial for response planning as it facilitates broad situational awareness of critical infrastructure status such as building damage, flooding, and obstructions to access routes. Damage mapping at this scale would require hundreds of expert person-hours. However, a combination of crowdsourcing and recent advances in deep learning reduces the effort needed to just a few hours in real time. Asking volunteers to place point marks, as opposed to shapes of actual damaged areas, significantly decreases the required analysis time for response during the disaster. However, different volunteers may be inconsistent in their marking. This work presents methods for aggregating potentially inconsistent damage marks to train a neural network damage detector.",
author = "Danil Kuzin and Brooke Simmons and Olga Isupova and Steven Reece",
note = "3rd Workshop on Artificial Intelligence for Humanitarian Assistance and Disaster Response (NeurIPS 2021); NeurIPS 2021 : Thirty-fifth Conference on Neural Information Processing Systems ; Conference date: 06-12-2021 Through 14-12-2021",
year = "2021",
month = dec,
day = "12",
language = "English",
url = "https://nips.cc/",

}

RIS

TY - CONF

T1 - Disaster mapping from satellites

T2 - NeurIPS 2021

AU - Kuzin, Danil

AU - Simmons, Brooke

AU - Isupova, Olga

AU - Reece, Steven

N1 - Conference code: 35th

PY - 2021/12/12

Y1 - 2021/12/12

N2 - High-resolution satellite imagery available immediately after disaster events is crucial for response planning as it facilitates broad situational awareness of critical infrastructure status such as building damage, flooding, and obstructions to access routes. Damage mapping at this scale would require hundreds of expert person-hours. However, a combination of crowdsourcing and recent advances in deep learning reduces the effort needed to just a few hours in real time. Asking volunteers to place point marks, as opposed to shapes of actual damaged areas, significantly decreases the required analysis time for response during the disaster. However, different volunteers may be inconsistent in their marking. This work presents methods for aggregating potentially inconsistent damage marks to train a neural network damage detector.

AB - High-resolution satellite imagery available immediately after disaster events is crucial for response planning as it facilitates broad situational awareness of critical infrastructure status such as building damage, flooding, and obstructions to access routes. Damage mapping at this scale would require hundreds of expert person-hours. However, a combination of crowdsourcing and recent advances in deep learning reduces the effort needed to just a few hours in real time. Asking volunteers to place point marks, as opposed to shapes of actual damaged areas, significantly decreases the required analysis time for response during the disaster. However, different volunteers may be inconsistent in their marking. This work presents methods for aggregating potentially inconsistent damage marks to train a neural network damage detector.

M3 - Conference paper

Y2 - 6 December 2021 through 14 December 2021

ER -