Google’s DeepMind team has unveiled ImageFX, a new text-to-image AI tool. It uses transformer models to generate images from textual prompts. Notably, the tool operates on a higher level of abstraction than its predecessors – it focuses more on overall structure and less on specifics. This allows images to be manipulated without requiring specific photo-editing software. The technology could have potential uses in video games and movies.
https://www.engadget.com/google-reveals-another-text-to-image-generative-ai-tool-imagefx-150659065.html?src=rss