Nvidia’s next AI can turn words into photorealistic images

DMCA / Correction Notice
- Advertisement -


Nvidia deep learning technologies continue to do amazing and strange things. Just a few weeks ago we saw how the company can use AI to automatically match voice lines to 3D animated faces. It’s the cool kind of technology that can help people create great things with ease, or in the case of Nvidia’s latest unveiling, potentially awesome things, but with ease nonetheless.

- Advertisement -

we saw for the first time Nvidia’s Gaugan a few years ago, It was demonstrated to convert basic doodles into photorealistic images with a click of a few buttons. It’s a very neat thing, and definitely worth playing with. Now GuaGAN2 is out And it doesn’t even need your sketches to create highly detailed landscape images.

Instead, deep learning models can transform text into surprisingly complex images. It’s been trained on 10 million landscape images, so that’s what it seems to be doing. The example shows how the image changes and becomes more in line with the text, meaning changes are easily made on the fly.

black friday deals
advertisement

- Advertisement -

Black Friday 2021 Deals: The Place To Go For All The Best Early Black Friday Bargains.

GuaGAN 2 works by using the technology we saw at GauGAN that can convert shapes into images with text-to-image generation. This means that a mixture of pictures and words can be used to create these images. Users can start with a general detail to get a comprehensive image and are then able to try and change smaller details by hand.

If you want to play with some of this technique, you can try this Nvidia’s interactive AI demo, however we had trouble getting it to work. Or, if you have an Nvidia RTX GPU, you can download nvidia canvas for free, It offers a good taste of painting with AI, which is impressive to see the work done with your own sketches in real time.

insert quote here

As to the implications of this technology in games, deep learning is likely to reduce time for developers from basically anything. People can quickly mock up test scenes in seconds with the technology. Haven’t figured out what the lookout looks like from your alien world rock, need a reference image you can’t find online? It has lots of practical uses for anyone looking to create.

Although personally I’m hoping someone will use this technique in a cool pseudo text adventure where you write scenes in life. It could also work with Nvidia’s planned AI haptics to be even more amazing. or disturbing.

- Advertisement -

Stay on top - Get the daily news in your inbox

Recent Articles

Related Stories