They say a picture is worth a thousand words, but with NVIDIA’s new GauGAN2 artificial intelligence (AI), you can create a picture with just three to four words. The deep-learning model is able to generate a photorealistic image of different scenes using text, sentences, or phrases that you enter into the program.
For example, if you type “Highway surrounded by mountains” in the input text field, the AI will generate an image of a highway surrounded by mountains. You can change any keyword, eg. “desert” instead of “mountains”, and a new image will be created based on your text input. You can further edit and customize the image with the tools available on the program.
GauGAN2 is an advanced version of the previous GauGAN AI that turned rough sketches into landscape images. The technology was eventually used to develop the NVIDIA Canvas app.
Check out some of the image examples for GauGAN2 below.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Know more in the video below
Tutorial for using GauGAN2 web demo
• NVIDIA Research Blog Post
• More NVIDIA AI Demos
Share this post with a designer friend and voice your views in the comments below.