Highly Detailed Population Maps Created By ‘Machine Learning’

Facebook is a whole lot more than just poking one another in fun, or perhaps to post up photos of your latest vacation and outing as an effort to ‘market’ yourself. Just in case you were wondering, Facebook’s ‘machine learning’ is well capable of creating highly-detailed population maps as well. The social media giant also claims that the maps will be able to identify buildings and other man-made structures scattered across 20 countries with a 5-meter resolution, where this might be utilized to improve Internet access for those living in rural areas as well as for research purposes, such as improving responses in the event of a natural disaster.

AI (artificial intelligence) machine learning technologies which enable computers to figure out the way to recognize patterns as well as arrive at conclusions could see action in the translation of languages, recognizing objects in photos, as well as filter email. Facebook’s competence in computer vision is harnessed to conjure detailed maps of where people live, as well as factors like how population density are able to impact their ability to hook up to the Internet.

This is an effort that can be attributed to Facebook’s Connectivity Labs, and the computing ‘muscle’ behind this effort is enormous – we are talking about analyzing 350 terabytes worth of images which would cover 21.6 million square kilometers.

Edwin Kee
Categories: Computers
Tags: AI (Artificial Intelligence), artificial intelligence, Facebook