Building Damage Mapping with Self-PositiveUnlabeled Learning

11/04/2021

∙

Humanitarian organizations must have fast and reliable data to respond to disasters. Deep learning approaches are difficult to implement in real-world disasters because it might be challenging to collect ground truth data of the damage situation (training data) soon after the event. The implementation of recent self-paced positive-unlabeled learning (PU) is demonstrated in this work by successfully applying to building damage assessment with very limited labeled data and a large amount of unlabeled data. Self-PU learning is compared with the supervised baselines and traditional PU learning using different datasets collected from the 2011 Tohoku earthquake, the 2018 Palu tsunami, and the 2018 Hurricane Michael. By utilizing only a portion of labeled damaged samples, we show how models trained with self-PU techniques may achieve comparable performance as supervised learning.

READ FULL TEXT

Building Damage Mapping with Self-PositiveUnlabeled Learning

Self-Supervised Learning for Building Damage Assessment from Large-scale xBD Satellite Imagery Benchmark Datasets

Data Distillation: Towards Omni-Supervised Learning

Masked Self-Supervision for Remaining Useful Lifetime Prediction in Machine Tools

Harnessing Unlabeled Data to Improve Generalization of Biometric Gender and Age Classifiers

Self-Supervised Learning with Limited Labeled Data for Prostate Cancer Detection in High Frequency Ultrasound

Classification of structural building damage grades from multi-temporal photogrammetric point clouds using a machine learning model trained on virtual laser scanning data

Class Mean Vectors, Self Monitoring and Self Learning for Neural Classifiers

Building Damage Mapping with Self-PositiveUnlabeled Learning

Related Research

Self-Supervised Learning for Building Damage Assessment from Large-scale xBD Satellite Imagery Benchmark Datasets

Data Distillation: Towards Omni-Supervised Learning

Masked Self-Supervision for Remaining Useful Lifetime Prediction in Machine Tools

Harnessing Unlabeled Data to Improve Generalization of Biometric Gender and Age Classifiers

Self-Supervised Learning with Limited Labeled Data for Prostate Cancer Detection in High Frequency Ultrasound

Classification of structural building damage grades from multi-temporal photogrammetric point clouds using a machine learning model trained on virtual laser scanning data

Class Mean Vectors, Self Monitoring and Self Learning for Neural Classifiers