Few-shot learning for post-earthquake urban damage detection

Abstract

Among natural disasters, earthquakes are recorded to have the highest rates in human loss in the past 20 years. Their unexpected nature has severe consequences on both human lives and material infrastructure and demands urgent action. For effective emergency relief, it is necessary to gain awareness about the level of damage in the affected areas. The use of remotely sensed imagery is popular in damage assessment applications, however it requires a considerable amount of labeled data, which are not always easy to obtain. Taking into consideration the recent developments in the fields of Machine Learning and Computer Vision, this thesis investigates and employs several Few-Shot Learning (FSL) strategies in order to address data insufficiency and imbalance in post-earthquake urban damage classification. The contribution of this work is double: we manage to prove that oversampling is the most suitable data balancing method for training Deep Convolutional Neural Networks (CNN) when compared to cost-sensitive learning and undersampling, and to demonstrate the feasibility of Prototypical Networks in a damage classification problem.