Google’s next big AI-powered photo feature could be RealFill: Here’s how it will work

Google’s next big AI-powered photo feature could be RealFill: Here’s how it will work
HIGHLIGHTS

Google seems to be preparing to unveil its next AI-powered photo feature.

Google has filed a trademark for what's being referred to as RealFill technology.

RealFill will be able to significantly enhance the precision of image expansion.

Google has been actively pushing forward with generative AI, prominently showcasing its capabilities through features like Magic Editor. Now, the tech giant seems to be preparing to unveil its next major AI-powered photo feature named RealFill.

Last week, Google discreetly filed a trademark for what’s being referred to as RealFill technology. This filing was filed via the European Union Intellectual Property Office (EUIPO) and the US Patent and Trade Office (USPTO), reports Android Authority.

Also read: Google’s AI-powered ‘Speaking practice’ feature will enhance your English skills: Here’s how

Google's next big AI-powered photo feature could be RealFill: Here's how it will work
Credit: Android Authority

“Providing non-downloadable software using artificial intelligence (AI) for inpainting images; Providing online non-downloadable software for creating generative models,” reads the description of the trademark.

Also read: Google’s Circle to Search could soon offer another useful function besides web searches: Check details

Google's next big AI-powered photo feature could be RealFill: Here's how it will work
Credit: Android Authority

RealFill technology actually came to light late last year through a paper and website by a collaboration between researchers from Google and Cornell University. Titled “Reference-Driven Generation for Authentic Image Completion,” the paper details a method for enhancing the accuracy of image expansion and inpainting.

RealFill will be able to significantly enhance the precision of image expansion and inpainting. It achieves this by leveraging the reference of up to five images, ensuring a more thorough and detailed completion process.

Despite its promising capabilities, the RealFill technology does come with some limitations, as highlighted by the research team. One significant drawback is the requirement for a “gradient-based fine-tuning process” on input images, which tends to slow down the overall process. 

Additionally, if the reference images are very different from the target image, it might struggle to make a good final picture. Lastly, it might not work well with text. These challenges show that RealFill still needs some work to be perfect.

While filed patents or trademarks don’t guarantee RealFill will become a commercial product, it’s reasonable to speculate that it could be integrated into future Pixel series phones or Google Photos if Google plans to release it commercially.

Ayushi Jain

Ayushi Jain

Tech news writer by day, BGMI player by night. Combining my passion for tech and gaming to bring you the latest in both worlds. View Full Profile

Digit.in
Logo
Digit.in
Logo