Analysis of Deep Learning Techniques for Maasai Boma Mapping in Tanzania

Underdeveloped countries in sub-Saharan Africa often contain cultural subpopulations that are underserved in regard to health and education. This perpetuates the health challenges of the country as a whole, and it is therefore of interest to be able to automatically map the subpopulation for the health services delivery. International nonprofit health organizations have often taken the lead in these efforts, providing humanitarian aid (e.g., clean water and food) as well as health care. This is necessary, as the ethnic subpopulations are not well integrated into the society and the existing health care systems. In this study, we explicitly explore the Maasailand of Tanzania, to evaluate the use of deep neural networks (DNN) to aid in the automatic visual analysis of remote sensing data to geolocate Maasai boma structures. We investigate the performance of four state-of-the-art DNN as classifiers of boma presence within high-resolution imagery; all showing over 95&#x0025; F1 score performance. Additionally, we scan over 3900 km<inline-formula><tex-math notation="LaTeX">$^{2}$</tex-math></inline-formula> of high-resolution imagery, combining a ProxylessNAS with broad area aggregation and mapping techniques and demonstrate the discovery of hundreds of boma, many that were not discovered by human analysts performing visual scans. The trained ProxylessNAS model generates a classified vector response field (CVRF). The CVRF is aggregated by a mode-seeking algorithm to detect potential locations of boma structures within the study area. The model detected numerous human false negatives (HFNs) and achieved 94.022&#x0025; TPR and 95.395&#x0025; F1 score using an aggregation aperture of 250 m within a 76.620 square kilometers area of interest.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.