IDEABench: Benchmark Dataset for Mapping Deprived Urban Areas

A collaborative data collection platform (portal.ideatlas.eu/ ) was created to support community involvement and improve data quality. As a key outcome of this research, we are excited to announce the release of IDEABench , a groundbreaking benchmark dataset that combines multiple sources of data to help researchers improve their understanding of deprived urban areas. This benchmark dataset contains image patches from two satellite systems, Sentinel-1 and Sentinel-2, along with precomputed buildup density information.

Furthermore, it includes detailed annotations for three distinct categories: deprived urban areas, non-deprived urban areas, and non-built-up areas. The dataset comprises a total of 47,476 patches , each with a size of 128×128 pixels from 8 cities across the globe, including Nairobi, Medellín, Mumbai, Buenos Aires, Lagos, Jakarta, Mexico City, and Salvador. IDEABench has the potential to support a wide range of applications, from identifying areas of deprivation and tracking urban growth to informing policy decisions and promoting more equitable urban development.

We believe that this dataset will be a valuable resource for the research community and look forward to seeing the innovative solutions it will enable. The IDEABench dataset is an outcome of the IDEAtlas project, a research initiative funded by the European Space Agency (ESA). The project is led by the Faculty of Geo-Information Science and Earth Observation at the University of Twente in partnership with GeoVille Information Systems and Data Processing GmbH.