The tool uses a neural network that estimates how "far" each pixel in a photo is from the camera in order to create a "3D photo." The AI was trained on a large set of data, meaning it works on ...