Gemini will offer the option to edit artificial intelligence-generated images with its new fine-tuning feature

Although we have become accustomed to human images with more than ten fingers or deformed faces in artificial intelligence-generated images, artificial intelligence companies continue to develop their models to create more realistic images. Google Gemini is working on a new fine-tuning feature to overcome this problem. Firstly Android Authority The feature, noticed on an unfinished code, will allow users to make detailed edits to their images created by artificial intelligence.

Currently, Gemini’s text-to-image tools do not offer the option to edit the image after creating it. Instead, users are trying to address the issue by creating new images through new prompts. We should point out that trying to solve small errors in details this way is not very productive. Gemini’s fine-tuning feature will meet the need for changes in a limited scope with two editing methods.

The first of these; The option allows users to submit a prompt about an AI-generated image. Thus, users can request changes to only one feature instead of changing the entire visual. For example, you may want to change the background of a created image while preserving the objects in it.

The second method in the code is a more interactive approach that resembles the Circle to search feature. Users will be able to use their fingers or stylus to circle the part of the image they want to change. Once the field is selected, users will be able to define the desired changes. So Gemini will understand that the instructions only relate to the circled part.

In fact, many AI image generators offer similar features, including ChatGPT with DALL-E integration, as well as Ideogram.ai and Adobe Firefly. Google’s plan to include these fine-tuning tools represents a technical leap for Gemini. It is not yet known when the feature will be available to users.