AI and ML models are particularly great at recognizing the image features that have been exposed. Schools are like other building infrustructures, and they have their primary purpose. It provides functions like public gathering, public recreation, shelter and even polling stations. Therefore, schools may have unique features that other buildings don’t have.From overhead imagery they can show as U, O, I, H shapes, they have basketball, playground, swimming pool. The turn to have a cluster of buildings that have some roof color. The building size is bigger compared to surrounding residential buildings. AI models can be trained to recognize school buildings very well. At the same time, we can utilize cloud computing and modern deep learning techniques to speed up model training and inference that can scan and search for schools in a really fast manner.

However, distinguished school features that have been feature engineered to train the model also means we may introduce human bias to the model. In the end the model may be able to recognize schools that are in distinguished building complexes, similar buildings rooftop, having swimming pools or basketball courts, but really bad at recognizing schools that have smaller building size and in poorer neighborhoods or even densely populated urban areas.