We introduce OpenEarthMap, a benchmark dataset, for global high-resolution land cover mapping. OpenEarthMap consists of 2.2 million segments of 5000 aerial and satellite images covering 97 regions from 44 countries across 6 continents, with manually annotated 8-class land cover labels at a 0.25--0.5m ground sampling distance. Semantic segmentation models trained on the OpenEarthMap generalize worldwide and can be used as off-the-shelf models in a variety of applications. We evaluate the performance of state-of-the-art methods for unsupervised domain adaptation and present challenging problem settings suitable for further technical development. We also investigate lightweight models using automated neural architecture search for limited computational resources and fast mapping. The dataset is available at https://open-earth-map.org.

The government plans to utilize artificial intelligence to constantly observe and analyze images of the Earth's surface and location data provided by satellites, mainly for disaster prevention, sources said Tuesday. Through the AI-based analysis, the internal affairs ministry aims to predict the risk of landslides by observing steep slopes, among other measures, hoping that the project will lead to the creation of new services by businesses and local governments. On Thursday, the ministry will launch an expert panel to discuss details of the project. The panel is expected to draw up proposals as early as June. Many businesses have advised the ministry that AI-based analysis of satellite data will be particularly effective in developing new disaster reduction services, including forecasts of tsunami arrival times for other countries, informed sources said.