Artificial intelligence, which successfully transforms text into visuals, has recently turned social media upside down. There is a new competitor in this market led by DALL-E 2 and Midjourney developed by OpenAI. Google researchers announced the artificial intelligence called DreamFusion. harunmudak
Google AI DreamFusion turns text into images
Google has actually been in the text-to-image creation market since last year. DreamFusion has been reintroduced as the latest update to Dream Fields, a text-to-text rendering tool that was first introduced in 2021.
Created using Neural Radiance Field (NeRF), artificial intelligence combines a highly advanced neural network to create 3D images. According to Google’s statement, users will be able to turn 2D texts into 3D with DreamFusion, which is still training. As you can imagine, each transaction will strengthen artificial intelligence.
Ben Poole, research scientist and project director at Google Brain, said, “We are pleased to announce DreamFusion, our new method for text-to-3D. We optimize NeRF from the ground up using a pre-trained text-to-image model. There is no need for 3D data”.
It is possible to say that DreamFusion has created very successful visuals. According to the statement on the website, the rendered images will be consistent with “high-quality normal, surface geometry and depth.” OpenAI has been developing an artificial intelligence algorithm called Dall-E 2 for a while. In fact, the Dall-E Mini version of artificial intelligence has recently gone viral on social media. Dall-E, who managed to convert the texts into images, may even have created his own dictionary to understand some words.