In keeping with a publish shared by Dylan Roussel on social media platform X (noticed by 9to5Google), an unreleased Google Bard changelog (dated January 18) has revealed how customers can “Create pictures with Bard.”
Creating pictures with Google Bard: The way it may fit
Like different instruments, customers will be capable of “merely enter a number of phrases to convey your creativeness to life.” Google may even permit customers to obtain or “Generate extra” pictures from the immediate they supply.The picture technology software in Bard is predicted to be powered by Google’s Imagen household of fashions.
Google introduced Imagen 2 with “considerably improved picture high quality” in December. This skill to create “high-quality, photorealistic, high-resolution, aesthetically pleasing pictures” is aimed toward enterprise use instances. This software can be utilized for brand technology or creating pictures with textual content overlays.
In October, Google Search launched an analogous picture generator. Customers who’re enrolled in SGE this system can discover this software at google.com/search/pictures. Google permits customers to generate “Draw” or “sketch” pictures from prompts straight entered into the Search area. SGE generates 4 pictures at a time. Google additionally has each day limits on picture technology with this software
Google has confirmed that SGE doesn’t generate pictures of dangerous, deceptive, or specific content material. Furthermore, prompts with notable folks or photorealistic faces are additionally blocked. Other than this, each picture generated has metadata labelling and embedded/invisible watermarking (utilizing Google’s SynthID).
Google Bard’s picture generator can be anticipated to supply comparable security options.This service joins user-facing picture technology capabilities in Google Slides and Meet (for customized backgrounds) as a part of Duet AI.