A new AI-based image editing feature has been introduced by Google to its ‘Gemini (Gemini)’ service, as announced on their official blog on April 30th (local time). It is available on both iPhone and Android apps, allowing anyone to easily alter photos just by entering a sentence.
Previously offered only on a separate platform called ‘AI Studio,’ this feature can now be directly utilized in the Gemini app. Users can upload photos to the app and input sentences like “Change the hair to blonde” or “Switch the background to a night sky,” allowing for the automatic editing of the background, colors, and character expressions.
This feature is noteworthy because AI automatically adjusts detailed elements such as color tone, lighting, and emotional expression without complicated tools. The maximum editable size is 1024×1024, with larger images expandable using external tools.
The editing process is cumulative, relying on a flow of conversation rather than single, isolated commands. Consequently, edits seamlessly build upon previous instructions given by the user.
The edited images have digital watermark technology called ‘SynthID’ developed by Google DeepMind, which is automatically embedded. It is invisible to the naked eye, but pixels include information to track it as AI-generated. The watermark remains even after cropping, filtering, or compressing and can be detected with specialized tools.
Google is offering this function progressively in 45 languages, including Korean. It is expected that domestic users will soon be able to use this feature in the Gemini app.
Google stated, “This will become a tool that enables anyone to create and edit photos easily without intricate technology.”