Google has announced significant updates to its Gemini platform, introducing Custom Gems and enhanced image generation capabilities with Imagen 3. According to blog.google, these advancements promise to deliver a more refined user experience while adhering to strict safety and design principles.

Enhanced Image Generation with Imagen 3

Imagen 3 brings advanced image generation capabilities that come with built-in safeguards. The new model adheres to Google’s product design principles and performs favorably across a wide range of benchmarks compared to other image generation models available. Similar to its predecessor, Imagen 2, the new model utilizes SynthID, a tool for watermarking AI-generated images to ensure authenticity and traceability.

Google emphasizes user control throughout the creative process. If the initial image does not meet expectations, users can provide feedback to Gemini, which will generate a new image based on the input. This iterative approach aims to enhance user satisfaction and creative flexibility.

Rollout of Image Generation of People

Over the coming days, Google will roll out the capability to generate images of people, initially available in an early access version for Gemini Advanced, Business, and Enterprise users. This feature will be released in English first, with plans to expand to more users and languages soon.

Technical improvements have been made to enhance the product, including refined evaluation sets, red-teaming exercises, and clear product principles. However, Google has set boundaries for this technology, stating that it does not support the generation of photorealistic, identifiable individuals, depictions of minors, or excessively gory, violent, or sexual scenes.

User Feedback and Continuous Improvement

Google acknowledges that not every image generated by Gemini will be perfect. The company is committed to listening to user feedback from early adopters to continuously improve the platform. By gradually rolling out these features, Google aims to refine the user experience and bring the new capabilities to a broader audience.

To be among the first to experience these new features, users are encouraged to try Gemini Advanced or sign up for Gemini for Workspace.

Image source: Shutterstock