Saturday, July 2, 2022
Home Insider Google takes on OpenAI with flashy text-to-image generator

Google takes on OpenAI with flashy text-to-image generator


The AI imagery competition is getting personal.

Google this week unveiled a new challenger to OpenAI’s vaunted DALLE-2 text-to-image generator — and took shots at its rival’s efforts.

Both models convert text prompts into pictures. But Google’s researchers claim their system provides “unprecedented photorealism and deep language understanding.”

Greetings humanoids

Subscribe now for a weekly recap of our favorite AI stories

: Example qualitative comparisons between Imagen and DALL-E 2 [54] on DrawBench prompts from Conflicting category. We observe that both DALL-E 2 and Imagen struggle generating well aligned images for this category. However, Imagen often generates some well aligned samples, e.g. “A panda making latte art.”