Towards Sustainable Census Independent Population Estimation in Mozambique
This work addresses the need for reliable population data for policy-making in vaccination and infrastructure planning in Mozambique, but it is incremental as it builds on existing census-independent methods with specific enhancements.
The study tackled the problem of intercensal population estimation in Mozambique by using remote sensing and microcensus data, finding that predictions improved when incorporating building footprint area estimated through transfer learning and additional annotations compared to using only publicly available features.
Reliable and frequent population estimation is key for making policies around vaccination and planning infrastructure delivery. Since censuses lack the spatio-temporal resolution required for these tasks, census-independent approaches, using remote sensing and microcensus data, have become popular. We estimate intercensal population count in two pilot districts in Mozambique. To encourage sustainability, we assess the feasibility of using publicly available datasets to estimate population. We also explore transfer learning with existing annotated datasets for predicting building footprints, and training with additional `dot' annotations from regions of interest to enhance these estimations. We observe that population predictions improve when using footprint area estimated with this approach versus only publicly available features.