Using Artificial Neural Network Models to Assess Hurricane Damage through Transfer Learning

Abstract

Coastal hazard events such as hurricanes pose a significant threat to coastal communities. Disaster relief is essential to mitigating damage from these catastrophes; therefore, accurate and efficient damage assessment is key to evaluating the extent of damage inflicted on coastal cities and structures. Historically, this process has been carried out by human task forces that manually take post-disaster images and identify the damaged areas. While this method has been well established, current digital tools used for computer vision tasks such as artificial intelligence and machine learning put forth a more efficient and reliable method for assessing post-disaster damage. Using transfer learning on three advanced neural networks, ResNet, MobileNet, and EfficientNet, we applied techniques for damage classification and damaged object detection to our post-hurricane image dataset comprised of damaged buildings from the coastal region of the southeastern United States. Our dataset included 1000 images for the classification model with a binary classification structure containing classes of floods and non-floods and 800 images for the object detection model with four damaged object classes damaged roof, damaged wall, flood damage, and structural damage. Our damage classification model achieved 76% overall accuracy for ResNet and 87% overall accuracy for MobileNet. The F1 score for MobileNet was also 9% higher than the F1 score of ResNet at 0.88. Our damaged object detection model achieved predominant predictions of the four damaged object classes, with MobileNet attaining the highest overall confidence score of 97.58% in its predictions. The object detection results highlight the model’s ability to successfully identify damaged areas of buildings and structures from images in a time span of seconds, which is necessary for more efficient damage assessment. Thus, we show that this level of accuracy for our damage assessment using artificial intelligence is akin to the accuracy of manual damage assessments while also completing the assessment in a drastically shorter time span.

On Transfer Learning for Building Damage Assessment from Satellite Imagery in Emergency Contexts

“…For example, ref. [37] studied the use of well-established convolutional neural networks and transfer learning techniques to predict building damages in the specific case of hurricane events. The model's ability to transfer to a future disaster was first studied by Xu et al [38].…”

(Retracted) Hurricane damage assessment in satellite images using hybrid VGG16 model

“…In the case of classification, ResNet achieved an accuracy of 87% and for object detection, a confidence score of 97.58% was achieved. 15 ML models have also been used for making predictions about hurricanes. In order to protect the lives of people, predictions of hurricanes need to be done.…”

“…Extreme gradient boosting is an ensemble model that is made up of many base learners. Base learners are generally obtained from the training data by the base learning models that could be a decision tree or other ML algorithms: 37,38 E Q -T A R G E T ; t e m p : i n t r a l i n k -; e 0 1 5 ; 1 1 6 ; 4 9 5 F ¼ ff1; f2; f3; f4; : : : ; fmg (15) is the set of base learners.…”

“…One thousand images have been used in this study for classification purposes, and 800 images have been used for object detection. In the case of classification, ResNet achieved an accuracy of 87% and for object detection, a confidence score of 97.58% was achieved 15 …”

A case study and parametric analysis of predicting hurricane-induced building damage using data-driven machine learning approach

“…For decades, ML has been applied to wind engineering subfields such as predicting windstorm intensity and frequency, incorporating topographic and aerodynamic features into wind models such as those in computational fluid dynamics, and as surrogate models to mitigate the expense of complex computational models (Wu and Snaiki, 2022). In recent practice, ML is utilized in the reactive categorization of building damage after hurricane impact by comparing preand post-storm imagery (e.g., Li et al, 2019;Calton and Wei, 2022;Kaur et al, 2022) and for near real-time detection of damage via analysis of social media posts (e.g., Hao and Wang, 2019;Yuan and Liu, 2020). These methods can be valuable for prioritizing emergency response allocation in the early hours following a hurricane.…”