Seeing is believing, but NVIDIA can boast of showing off new AI that can transform a simple sketch into a masterpiece of photorealism. It is a deep learning model developed by NVIDIA Research that manages to do all the magic.
The tool takes advantage of called them as GANs (adverse generative networks) to turn those sketches into fully realistic images. This interactive app that uses the aforementioned model has been baptized as GauGAN.
GauGAN could offer a powerful tool by creating virtual worlds to everyone. From what could be architects, game designers or any professional who needs to create in some way a type of multimedia content.
In other words, we are talking about an artificial intelligence that understand how the real world is generated before our eyes. So it becomes a perfect tool to simply create quick environments with ideas that come up in no time.
Bryan Catanzaro, Vice President of Deep Learning Research at NVIDIA, maintains that the technology behind GauGAN is like a smart brush that can fill in those sketches with details that anyone can quickly draw without much knowledge in drawing.
A model of deep learning trained with millions of images and that in a matter of seconds it is capable of filling an abstract shape with trees, rocks, reflections and all kinds of elements with which to create a realistic scene. It is as if we draw a quick sketch and it is automatically filled with a perfect panorama with its trees, sun, sky ...
Polenta magic artificial intelligence, as it happens with this website that generates random people's faces, who works as an artist when drawing or painting something. Now we have to wait when this tool is online to be able to play with it and see it in situ. We will announce it in these parts at that time.
I love her