fbpx
News

Facebook aims to use AI-generated maps to help global crisis relief

Facebook app on iPhone

Facebook has revealed new information on a long-gestating project that leverages artificial intelligence to create accurate maps that can be used in humanitarian efforts.

In 2016, Facebook revealed high-resolution maps for 22 countries that were generated using AI as part of its Population Density Maps project.

Now, the company is expanding on that original lineup to include maps that cover the “majority” of the continent of Africa. The ultimate goal of the project is to “eventually map nearly the whole world’s population, according to Facebook.

As Facebook notes, precision is key in mapping, which isn’t always easy to maintain when it is done manually. With roughly 1.3 billion people living across Africa, it can be difficult and tedious to narrow down exactly how many are living in specific regions, cities and towns. While satellite imagery can help, humans still have to label each structure and then cross-reference these images with census data.

With that in mind, Facebook is automating the mapping process using data from open-source mapping project Open Street Map. Using Open Street Map, Facebook engineers can train a computer to recognize buildings in satellite imagery. By identifying all of the structures, these engineers can then ignore fields, forests and other unoccupied land, thus saving a great deal of time while mapping.

Facebook says it will offer these maps for free, noting that this data can be invaluable in disaster relief and vaccination efforts.

However, Facebook is looking to benefit from greatly from these maps as well. As noted by The Verge, the company first unveiled this mapping project in 2016 as a means to “connect the unconnected” and get more people online and signed up to its social network.

Source: Facebook

MobileSyrup may earn a commission from purchases made via our links, which helps fund the journalism we provide free on our website. These links do not influence our editorial content. Support us here.

Related Articles

Comments