Google is enhancing its AI capabilities with a new precision image editing feature for Gemini, currently being tested in the latest Android beta. This upgrade aims to give users more control over AI-generated images, potentially rivaling competitors like ChatGPT's DALL-E.
New Editing Tools in Beta
The latest Google beta app for Android (version 15.40.31.29) introduces a more intuitive way to edit AI-generated images within Gemini. Users can now make selections for precise edits, allowing for targeted modifications to specific parts of an image. This feature addresses a key limitation in the current version, where users rely solely on text prompts for refinements.
How It Works
- Generate an initial image using text prompts
- Select specific portions of the image for editing
- Provide follow-up prompts for narrower, more focused changes
A demonstration video showcases the ability to generate an image of a dog, then make specific edits like changing the type of hat the dog is wearing.
Current Limitations
While promising, the feature is still a work in progress:
- Edits are not always precise or reliable
- Simple changes can sometimes alter unintended parts of the image
- Each edit takes time to process (over 10 seconds in some cases)
Competitive Landscape
This update positions Google Gemini to compete more directly with other AI image generation tools:
- ChatGPT's DALL-E
- Midjourney (currently considered a top AI image generator)
- Apple's upcoming Image Playground
Looking Ahead
As AI image generation tools rapidly evolve, Google's focus on precision editing could give Gemini an edge. However, the true impact of this feature will only be clear once it's officially released and thoroughly tested by users.
For now, those interested in trying out these new capabilities will need to wait for Google to roll out the feature more widely, which is expected to happen in the near future.