Partimap

Training satellite imagery classification algorithms to identify the different quality of south African slums with the help of their inhabitants

partimap training algorithm

AI technologies can efficiently handle complex tasks quickly. They can be replicated and assessed. One application is identifying slum expansion in countries where informal settlements are poorly recognized. While some current technologies can identify these settlements with accuracy, assessing their quality is essential. Current methods involve morphometric analysis, but we must delve deeper to understand the perceptions of slum inhabitants regarding different qualities.

We’ve conducted an experiment similar to Mercè and Arturo, training an algorithm to assess urban environment quality with input from slum inhabitants. The resulting dataset was used to evaluate the quality of various slums in South African cities.

Images

Promoters


300.000 Km/s

Publications

Related expertises

services

toolbox

topics