GauGAN2 by NVIDIA

GauGAN2 is a GAN (Generative Adversarial Network) model that merges various modalities such as text, semantic segmentation, sketches and styles into a single framework. This allows users to take an artist's concept and turn it into a high-quality AI-generated image.

Using GauGAN2, users can input a short description to quickly create an image's key features and theme. As an example, NVIDIA gives the phrase of a snow-covered mountain range, which can then be altered with drawings to make a particular mountain higher, add a few trees, or other alterations in the foreground, or clouds in the sky.

Although GAUGAN2 is still in its early stages of development, with the help of a limited data set, it has already shown potential to revolutionize asset creation in movies and games with its photorealistic landscapes and objects created with very little user input.

Sources:

• https://analyticsindiamag.com/nvidias-gaugan2-vs-openais-glide/
• https://www.notebookcheck.net/NVIDIA-s-AI-based-GAUGAN-2-tool-generates-Van-Goghesque-landscapes-from-words-and-phrases-you-input.581454.0.html

Discaimer: GauGAN2 doesn't use GPT-3