Facebook has made it possible to create 3D photos for the majority of iOS and Andriod device users, according to a company blog.
The tool was launched in 2018, but it could only be used on smartphones with two cameras.
According to the social network, you can make a three-dimensional picture from “almost any” standard 2D image.
- The company has trained a convolutional neural network on millions of publicly available 3D images and their depth maps. The network estimates the distance from the camera to each pixel in the image.
- The architecture of the network makes it possible to build a 3D image on mobile devices of even low power “in seconds,” writes Facebook.