Gemini Imagen 3 released: Google's new image generation model
Google has recently announced the re-introduction of Gemini's image creation capability with Imagen 3. After temporarily halting the feature earlier this year, the tech giant is now enabling users to utilize AI models for image generation. This is not the first time Google's Gemini platform has possessed such capabilities. The earlier version of Imagen faced backlash due to issues such as biased and inaccurate historical depictions, prompting Google to make necessary improvements.
Improvements in Imagen 3
In a blog post by Dave Citron, the senior director of Gemini, it was revealed that Imagen 3 has undergone significant upgrades in its image generation capabilities. The new version will be gradually rolled out to all users in the coming days, offering enhanced creative features and multilingual support. According to Google, Imagen 3 establishes a new benchmark in image quality, allowing users to generate images with minimal input.
Citron emphasized that Google has made substantial progress in depicting individuals through Imagen 3, incorporating technical enhancements, evaluation sets, red-teaming exercises, and clear product principles to ensure product integrity. The company acknowledges that there might be imperfections in the creative output of Imagen 3 but is committed to gathering user feedback for further enhancements.
Users are cautioned against creating photorealistic imagery, identifiable individuals, depictions of minors, or explicit content, reinforcing Google's commitment to responsible AI usage. The rollout of Imagen 3 will be gradual, with an emphasis on user feedback and continuous improvement of the feature.
Additional Features
Besides Imagen 3, Google has introduced other features to the Gemini platform. Advanced, Business, and Enterprise subscribers can now leverage Gems, a customized version of Gemini that enables users to create and interact with expert AI models.
Gems can be tailored to offer insights, refine strategies, or assist in various tasks, catering to diverse user needs. This new feature allows users to form a team of expert Gems to aid in project planning, event ideation, social media content creation, or task automation. Gems are equipped to remember detailed instructions, streamlining tedious or complex tasks for users across desktop and mobile devices.
Currently, Gems are being rolled out to Gemini Advanced, Business, and Enterprise users in over 150 countries, supporting multiple languages to enhance user experience and productivity.
With the release of Imagen 3 and the introduction of Gems, Google continues to innovate its Gemini platform, offering enhanced creative tools and AI capabilities to users worldwide.
