Dataset for Land/Water Semantic Segmentation in Tonga and other Pacific regions

Dataset used for training U-Net model for land/water semantic segmentation of Sentinel-2 images. Features 424 Sentinel-2 images of size 256x256 pixels, with corresponding targets (binary segmentation maps). Weight maps are included for training using a weighted loss function, in order to improve segmentation on important features, specifically small volcanic islands. The models' weights of the pre-trained models used for our research are included as .keras files. Code snippets include essential functionality to load the models, and pre-process Earth Engine images for input.