Is Thierry Ardisson’s program the television of tomorrow or yesterday?

What if meeting Dalida, Jean Gabin or Lady Di were possible in 2022? We would dream it, Thierry Ardisson has (almost) done it. In Time hotel, the first edition of which will air on Monday 2 May on France 3, the star presenter interviews deceased actors, singers and politicians. How is it possible ? Using one of the applications of artificial intelligence: deepfakes. More precisely, the presenter and her team have relied on the technologies of two French companies, Ircam Amplify and Mac Guff, to recreate, with one, the voices of the missing stars, and with the other, their faces.

Behind the algorithmic experiment, two French companies

“One of our areas of work is the voice in the broad sense,” explains Nicolas Pingnelain, commercial director of Ircam Amplify. Created in 2020, the company aims to economically promote the fundamental research carried out at the Acoustic / Music Research and Coordination Institute (Ircam). “Upstream, this means detecting emotions, markers of stress or anxiety, downstream, synthesizing a voice capable of reproducing this type of variation. The human being is even more sensitive to the voice than to the image, he continues, a nonsense allows us to detect if it is false. No wonder Thierry Ardisson’s team turned to French researchers: a monotone voice like that of our intelligent assistants to deal with Dalida’s intonations would never have worked.

Also for the faces it was necessary to create the illusion of reality. The team turned to Mac Guff, a 35-year-old French studio, creator of, among others Me, ugly and bad. “We have always had a research and development dimension, which allows us to remain at the forefront”, says its president Rodolphe Chabrier. We were therefore among the first to do morphing, relief… ”Five years ago, the entrepreneur embarked on artificial intelligence. Mac Guff soon created two tools: the Talking Picture, to make still images “talk”, and the Face Engine, dedicated to recreating faces in 3D. Both built on deep learning models, these technologies made it possible to model the faces of past stars on the bodies of the actresses and actors who gave Thierry Ardisson the replica during the filming of the docu-fiction.

How to raise the dead?

Why Thierry Ardisson is enthusiastic about it: Time hotel it is a documentary work. And it’s true: to train the algorithmic machines that reconstruct faces and voices, they need to be equipped with equipment. These training datasets consisted of every possible archive on each of the celebrities the show brings to life. “Television appearances, images given by families, films in the case of actors like Jean Gabin, enumerates Rodolphe Chabrier… Let’s take everything we find! “For the voice it is a bit simpler, explains Nicolas Pingnelain:” The important thing is to have samples of all the phonemes, of all the syllables that exist in the language. After, with a few tens of minutes of intervention, we can already have a good result. “As for the words spoken by Dalida, Lady Di or Coluche on the show, Thierry Ardisson points out, they were all really said in the lifetime of celebrities.

The result works quite well … But still, we can’t help but feel uncomfortable. Could it be the Valley of the Uncanny effect, which makes robots trying to mimic humans a little unpleasant to the person watching them? Is it the fear of a certain leveling of the faces, when one dwells on the slightly too fine nose of the artificial Dalida, on the photos of the shooting? “We had Dalida’s voice validated by her brother Orlando just two weeks ago and he himself cannot distinguish between the synthesized version and the real one”, assures Nicolas Pingnelain. “When they saw the result, the Gabin children found their father’s expressions, they were amazed! adds Rodolphe Chabrier. The production also made the decision to “blow up” the stars, he points out, “to look for the ‘iconic’ Dalida, as Thierry puts it.” In this case, the discomfort may stem from ethical issues: is it acceptable to bring dead people back to life? Response of all who have worked time hotel : not only is there no law that prohibits it, but those entitled to it have reviewed everything, the scripts, the faces, the voices, and everything has been validated.

Hotel Innovation?

“An instrument has no ideology, underlines Thierry Ardisson, this point is very important in the face of the a priori that can exist”. Rodolphe Chabrier conveys the same message and we can’t help but smile: in technology, the argument of neutrality is a classic. In the world of artificial intelligence, however, more and more researchers and activists believe that posing the problem in terms of good or bad use is not always relevant: in some cases, it would be simply necessary to refrain to develop a tool. For time hotelThat said, it is clear that Thierry Ardisson follows an idea that has haunted him for a long time: in 1994 he “resurrected” John Lennon in Gone with time. In 2002, it was Victor Hugo who invited on the set ofWe will have seen it all, along with Renaud and Guy Bedos, alive and well. Replacing lookalikes with deepfakes, maybe it’s just keeping up with the times?

This would mean losing all the innovation dimension required by the creation of the program: Rodolphe Chabrier as Nicolas Pingnelain underline the innovative aspect of the program and the work it requires. “60 minutes of faces recomposed on a 90-minute program, which has never been done before! exclaims the leader of Mac Guff, if only because the technologies accessible before the company built its algorithmic models would have made the operation too expensive. For text-to-speech, “our researchers modified the instrument less than a month ago and will likely do so again after the release oftime hotel says Nicolas Pingnelain: The show used applications straight from basic research. Any scientific advancement will improve the programs that have been built for the occasion.

A double cultural opportunity

In fact, maybetime hotel opens two cultural axes at the same time: the ability to get to know the interviewed stars better, first of all. After all, Thierry Ardisson’s stated goal is to “make culture spectacular and make knowledge captivating”. It is also an opportunity to see what a still young technology can give in the field of visual and sound special effects. Deep fakes have already been used in series – by Mac Guff, to rejuvenate Mathieu Amalric The office of legendsor the Russian operator Megafon, to play Bruce Willis in a commercial without him actually acting.

Difficult to predict if time hotel it will itself mark a turning point. On the other hand, working on this specific case has allowed Mac Guff developers and Ircam researchers to make progress in their respective fields. And among the applications they envision for their new algorithmic tools, Rodolphe Chabrier and Nicolas Pingnelain are sketching out various avenues such as voice authentication (to finally abandon passwords), dubbing (no more French voices that don’t correspond to American actors?) or, more classically, the retouching of faces and then of bodies. “With Face Engine you can recreate, rejuvenate, age, grow, transform someone into a burn victim in a very realistic way, for cinema or video games”, explains Rodolphe Chabrier. He who knows what show Thierry Ardisson will invent when it will also be possible to play on bodies and environments.

Leave a Reply