Labeled satellite imagery for training machine learning semantic segmentation models of coastal shorelines

A collection of data releases containing labeled satellite imagery for the purpose of training Machine Learning models to automate the task of shoreline extraction from satellite imagery. Shoreline mapping from satellite imagery, known as Satellite-Derived Shorelines or SDS, has the potential to transform coastal shoreline mapping for erosion hazard mapping and coastal resource assessment, among many potential uses. Automation of such tasks using Machine Learning is a crucial component of cost-saving and quality assurance for large-scale routine shoreline mapping. The associated image suitability dataset (https://doi.org/10.5066/P14MDKVJ) and semantic segmentations of satellite imagery (https://doi.org/10.5066/P1N4VI7H) are available.