Google prepared to bring back a function which creates images of people using its Gemini AI after pulling the plug on a previous version in February due to inaccurate results.
Dave Citron, a senior director of product at Google Gemini, wrote in a blog early access to the company’s new Imagen 3 generator will be available in English to Gemini advanced, business and enterprise users in the coming days.
During the initial launch, some elements of the image-generating feature in Gemini provided unintended results, which in some cases included culturally inappropriate visuals.
″We’ve worked to make technical improvements to the product, as well as improved evaluation sets, red-teaming exercises and clear product principles,” Citron stated.
He wrote Google made significant progress in providing a better user experience when generating images of people.
Citron stated Google does not “support the generation of photorealistic, identifiable individuals” or those depicting children, gore, violence or “sexual scenes”.
“Of course, as with any generative AI tool, not every image Gemini creates will be perfect, but we’ll continue to listen to feedback from early users as we keep improving,” he wrote.
“We’ll gradually roll this out, aiming to bring it to more users and languages soon”.
Comments