Disney Research Studios presented a new tool AI that ages or rejuvenates the face of the actors autonomously and precisely: the system FRAN (Face Re-Aging Network). All this thanks to a neural network that analyzes thousands of faces of artificially created “people”., to show more or less young faces in all possible expressions. Simplifying the work of make-up and special effects.
Disney, FRAN’s AI ages or rejuvenates the face of the actors
When the then 47-year-old Marlon Brando had to play the elderly progenitor of the Corleone family in The Godfather, he had to spending hours on makeup and keeping a denture in your mouth to deform the cheeks. Today an actor under fifty who wanted to play that role would need nothing more than an artificial intelligence. And one of the greatest talents in the history of cinema, but that’s another story.
Disney Research Studios has presented an academic essay on the AI FRAN system, because the applications of an artificial intelligence that ages and rejuvenates faces faithfully is interesting in many areas. But he also made a presentation video. Because, whether it comes from the skill of a make-up artist or the calculations of a machine, aging or rejuvenating a person is above all one of the magic of cinema.
Invented faces to age real faces
Like other neural networks, FRAN uses a database full of images to figure out how a face might naturally age or de-age. Many of these tools use data from volunteers or taken from public images. But for avoid privacy issues and any bias in the datathe image database contains pairs of randomly generated synthetic images in various ages.
This also helps researchers because it becomes difficult to have people re-propose the same photograph with the same expression years later, especially while also keeping the same light on stage. The only way without resorting to random generation is to ask volunteers to take pictures in the same poses as old shots. Not exactly a precise process.
The perfected version of an aging filter
Using this image database, FRAN takes the face of a real actor or actress and evaluates which wrinkles or details to add or remove, to automatically age or rejuvenate the person. Basically what aging filters do about Snapchatbut up enough detail to be good for the next Marvel movie or Star Wars.
Credit: Disney
However, Disney itself admits that there are limitations. For example, it cannot create images of actors and actresses when they were under twenty years of age, since the database does not take into account physical evolution during adolescence. Nor can it show thereceding of the forehead line or the onset of gray hair, since it was not expected in the data set.
This should reassure the makeup artists and special effects experts working in Hollywood. Disney doesn’t seem to have developed this tool to do without them, but rather to streamline the work process. Wigs and make-up will still be needed during filming, just as the video-editing experts will still have to work on the images in post-production. But the job should be easier. Or at least, we sincerely hope so.
Disney’s AI that ages and rejuvenates actors is perfect for franchises
The result looks in many cases better than seen with Robert Downey Jr in Capitan America, or more recently with Luke and Leia in the Star Wars series, at least visually. Even if some limits will remain: if the de-aging of Robert De Niro’s The Irish Man (on Netflix) it was well done, impossible to notice how the actor certainly didn’t move in his thirties.
However, Disney has every interest in developing an AI that ages and above all rejuvenates the actors. Franchises like the MCU have already taken advantage of young versions of Nick Fury (Samuel L. Jackson), Hank Pym (Michael Douglas), and Ego the Living Planet (Kurt Russell). As well as Leia (Carrie Fisher) e Wilhuff Tarkin (Peter Cushing) they have appeared in films not only rejuvenated, but literally brought back to life by special effects.
With more and more cinematic universes to manage, Disney might drastically reduce the cost of post-production of many of his works by focusing on artificial intelligence. Though fans are not always enthusiastic the result of these processes. However much technology advances, the heart of films always remains a good story well told: if we cheer for the characters on stage, even the rejuvenation of the AI will seem more realistic.