Google’s Gemini is again capable of generating people’s images. The company retracted this feature due to backlash over inaccuracies in generated images, particularly those depicting historical figures.
However, the search giant has started giving early access to Gemini Advanced subscribers. Google Workspace business and enterprise users will soon get this feature.
Google’s New AI Text-to-Image Generator
This new feature is backed by Imagen 3, Google’s latest version of AI text-to-image generator. It was launched earlier this month as part of AI Test Kitchen. It’s only available in English, but the company plans to expand it to more languages soon.
The Imagen 3 better understands natural language prompts, making it possible to create more realistic images with better detail and fewer mistakes than earlier versions.
It’s important to note that this feature differs from the new Reimagine feature introduced with the latest Pixel 9 series devices. With this feature, users can select specific objects or areas in a photo and modify them based on textual descriptions.
Dave Citron, Google’s senior director of product management for Gemini Experiences, highlighted that the new model incorporates significant technical improvements and safeguards. Users will still be restricted from generating images of identifiable individuals, minors, or excessively violent or sexual content. Citron emphasised that while not every generated image will be flawless, Google is committed to refining the tool based on user feedback.
However, Google will now have to face tough competition in this space. Rivals like OpenAI and Meta have already added similar capabilities to their AI chatbots.
The post Google Gemini Reintroduces AI-Generated People Feature appeared first on MySmartPrice.