NVIDIA’s GauGAN2 is a revolutionary GAN (Generative Adversarial Network) framework that allows users to quickly generate photorealistic images based on user-defined sketches and descriptions. This technology combines multiple modalities such as text, semantic segmentation, sketching and styling into one platform. For example, if you enter the phrase “snow capped mountain range” into GauGAN2 it will automatically create a landscape image featuring mountains with snow caps. You can then further customize your image by adding trees in the foreground or clouds in the sky. GauGAN2 is still early in development and has been trained using limited datasets but its potential for transforming asset creation for movies, games and other visuals is amazing. With just a few words of input from you, GauGAN2 can produce unique photorealistic landscapes and objects that look like they could have been painted by Vincent Van Gogh himself!