A new AI (Artificial Intelligence) tool will help you take a simple face scratch and turn it into a realistic photo.
The new algorithm take simple draws, analyze different aspects of the picture and create a full face image.
Scientists in the Chinese Science Academy in Beijing published a research and revealed an AI algorithm that take simple draws and turn them into a picture full with details.
The algorithm called DeepFaceDrawing, which help users with low drawing skills to produce high quality picture.
This is not the first time that AI is used to create pictures, a tool named Pix2Pix did it (very poorly) before, but the new AI tool looks much more promising than the old one.
The AI tool rely mostly on statistics, instead of demanding a full draw with all the elements, the tool go over all the objects in the data base, which includes faces and face parts, then it decides how to build the best face possible. So the tool analyze each face part first and then combine them all.
In the science paper the researches writes – “Recent deep image-to-image translation techniques allow fast generation of face images from freehand sketches. However, existing solutions tend to overfit to sketches, thus requiring professional sketches or even edge maps as input. To address this issue, our key idea is to implicitly model the shape space of plausible face images and synthesize a face image in this space to approximate an input sketch. Our method essentially uses input sketches as soft constraints and is thus able to produce high-quality face images even from rough and/or incomplete sketches.”
This is how the program works –