Showing what a movie character looked like in their youth—or might look like in old age—has always required significant effort and budget. Until recently, these transformations were labor-intensive and often far from perfect. But a research team at Disney has developed an innovative solution to this challenge.
Enter FRAN, a neural network whose name stands for **Face Re-Aging Network**. This AI-powered tool can change a character’s age on video in just seconds, according to Designboom.
A Magical Digital Assistant for Filmmakers
The Disney research team—including Gaspar Zoss, Prashanth Chandran, Eftychios Sifakis, Markus Gross, Paulo Gotardo, and Derek Bradley—describes FRAN as “the first practical, fully automated, and production-ready method for age modification in video footage.”
What once took days of post-production work can now be accomplished with a single click.
FRAN is trained on a massive dataset of synthetically generated faces across different ages. Using U-Net, a powerful image segmentation tool, the AI precisely predicts how each region of a face would age or rejuvenate, delivering impressively realistic results.
The tool can convincingly re-age faces within a range of 18 to 85 years old. The research team also released a demo video showcasing FRAN’s ability to age or de-age characters even while they move their heads or change facial expressions—something that has historically challenged visual effects teams.
Despite the promising results, Disney researchers acknowledge there’s still room for refinement. For instance, handling the graying of hair is one of the remaining technical hurdles.
Nonetheless, FRAN is poised to become a game-changing digital assistant for the film industry. It could soon eliminate the need for heavy prosthetic makeup and significantly cut costs associated with age-related visual effects in film production.