Fundamental to PARSEC has been the development of Deep Learning code to calculate wealth index for communities using open source Earth Observation and socio-economic data. The product, DeepWealth, is an end-to-end framework for computer-literate users skilled in Python and R to estimate and visualise well-being data. In the paper we describe its structure and show how it can be applied in different situations. The framework follows FAIR principles, providing test data, the source code, metadata and training checkpoints through github and zenodo, thus fulfilling the goals of reproducibility and replicability.
the paper
Ben Abbes, A., Machicao, J., CorrĂȘa, P.L.P., Specht, A., Devillers, R., Ometto, J.P., Kondo, Y., Mouillot, D., 2024. DeepWealth: A generalizable open-source deep learning framework using satellite images for well-being estimation. SoftwareX 27, 101785. https://doi.org/10.1016/j.softx.2024.101785
the code
Ben Abbes, A., Machicao, J., Correa, P.L.P., Specht, A., Devillers, R., Ometto, J.P., Kondo, Y., Mouillot, D., 2024. Source code: DeepWealth: A Generalizable Open-Source Deep Learning Framework using Satellite Images for Well-Being Estimation. https://doi.org/10.5281/zenodo.12189299