Google recently announced the upgrade of the Gemini AI chat assistant from Gemini 1.0 to the newer Gemini 1.5 Flash model. The latest model offers better speed and efficiency compared to the previously used model and also supports a 32k-long context window.
Now, a fresh APK Teardown report suggests that Google is working on saving time for the users who generate images using Gemini AI. Currently, to fine-tune an image generated using Gemini AI, the user needs to go back and create an image from scratch using the text prompts, including the minute details they wish to be included in the image.
This could take a lot of time if the output by Gemini isn"t perfect or in line with the text prompt entered by the user. So, Gemini lacks the feature to tweak an image after it has been generated. Thankfully, it may change soon.
In the latest Google beta app version 15.29.34.29 for Android, a new tool has been spotted that lets you fine-tune AI-generated imagery. This would allow users to tweak finer details on an image generated using Gemini that has already been created using the AI tool.
This would help the user in two ways. First, Gemini will be able to retain the core of an image, but certain things could be changed using natural text prompts. Second, Gemini will understand that you don"t need an entirely new image but want certain parts of the image tweaked.
Users will be able to fine-tune or tweak the details of the image using text prompts and manually circle the specific things they want to edit. The manual circling part could be pretty useful on phones with a stylus, as it would negate the need to enter text prompts to describe Gemini, which part of the image you are referring to.
Sadly, there is no time frame on when this feature will make its way to the stable version of the app.
Source and image: Android Authority