Facebook has a lot of pictures uploaded daily by hundreds of millions of people from all over the world, and for some of them he announced an interesting change that is possible thanks to an artificial intelligence system he developed.
Facebook announced that for the 3D pictures that people upload, a change was made that allows them to be published even if they are not recorded in portrait mode, so including pictures of other types, even 2D ones, can be transformed into 3D pictures.
Facebook it uses a neural network to calculate the distance between the camera and the subjects in the photos, for each pixel, and thus realizes what kind of depth must be applied so that even normal photos are 3D when they are viewed in its network.
Facebook it is not the first company from which we have a similar function, those from Google have been offering something similar for Pixel phones for some time, but in that case it was not possible to transform 2D photos into 3D photos, so the change is bigger than it seems.
“Our 3D Photos feature on Facebook launched in 2018 as an immersive new format for sharing images with friends and family. The feature was based on dual-lens "portrait mode" features only available on new, high-end smartphones. To bring this new visual format to more people, we've used state-of-the-art machine learning techniques to produce 3D photos from almost any standard 2D image. This system affects the 3D structure of any image, whether it's a new photo taken only on an Android or iOS device with a standard stock camera, or a decades-old image recently uploaded to a phone or laptop."
Facebook limited this functionality for iPhone 7 phones, or newer, mid-range Android phones that have a dual camera and were released in recent years, with some limitations, or Pixel series phones, so quite a lot of people can use it.
