How will Realfill change the game for AI image editing?

Realfill will use a technique called random masking for image editing and generation

Alexander Lewis - May 01, 2024

An undated image of a Realfill edit. — Realfill

Google, the tech giant, hints at a major innovation, 'RealFill' — an AI-generated method for image completion which fills in the missing areas of an image with details provided as a reference by the users.

Through a series of trademark applications and research findings, Google may be on the way to introducing this new AI model. This unique model will use the random masking technique for image editing and generation.

Google Realfill website was noticed online that operates using a reference image algorithm to perform image completion and painting features for creating the targeted image. It can generate new images effectively by understanding the subject and environment of the reference photos given by the users.

Google has filed trademark applications for its logo, hinting at introducing this feature as a commercial product. RealFill can enhance an image in various aspects, including technical features such as colour, size, and dimensions, as reported by Gadgets360.

Last year, Google released a handy feature called Best Take, which lets users combine the best parts of multiple group photos into one final image. RealFill looks like a more advanced version of this technology, and Google is planning to add it to future Pixel phones to enhance its camera capabilities.