.png)
Google is adding an artificial intelligence (AI)-powered native picture editing functionality to the Gemini app. This new feature, announced on Wednesday, allows users to make exact modifications to photographs without leaving the AI chatbot. This functionality was originally made available in March through the Google AI Studio platform. Using the Gemini 2.0 Flash AI model, users may ask the chatbot to rearrange components, add, replace, or remove items, enhance the backdrop, and more.
The Gemini app now supports native image editing
Gemini was quite late to the image creation game as compared to competitors like ChatGPT, Copilot, and Midjourney. While the function was initially released in early 2024, the Mountain View-based internet behemoth was forced to suspend and take back the service after multiple users discovered the chatbot was producing erroneous and insulting photos. Google then reintroduced image creation to Gemini in December 2024 with Gemini 2.0 Flash.
In a blog post, the tech titan revealed that the Gemini app will now support native picture editing via Gemini. The feature is being brought out gradually, and it may take a few days until all users have access to it. Notably, native picture editing will be available to consumers worldwide in over 45 languages.
The Gemini 2.0 Flash-powered capability was initially introduced in the AI Studio, but it is now more generally available. It will allow users to take an AI-generated image or upload a real shot and make conversational modifications to it. For example, a user can supply an image of a table and request that Gemini place some flowers on top of it.
According to the article, users may instruct Gemini to perform extensive alterations such as altering the backdrop, removing items, adding components, and even making tiny changes to the subject. An example of the latter would be to upload a photograph of yourself and request that the AI show you what you might look like with different hair colors.
One problem associated with this functionality is the proliferation of deepfakes. Google claims that its models are trained to decline any inappropriate or possibly damaging requests. Additionally, photos generated or altered using native image creation will carry the invisible SynthID digital watermark. The business is also testing the addition of a visible watermark to all AI-generated photographs.
#google #gemini #googlegemini #geminiapp #ai
