Disney Develops Facial Recognition Technology to Watch Audiences

“I always feel like somebody’s watching me…”

Miss Rosenby Miss Rosen
Photo: Courtesy of Disney Research

Disney Research takes film technology to the cutting edge with the development of a neural network that uses facial recognition software to reach the emotional reactions of hundreds of people sitting in the audience in order to determine whether or not the movie is evoking the emotions desired by filmmakers.

Also: Experts Predict When Artificial Intelligence Will Exceed Human Performance

In the new study released last month, researchers shared the findings of an experiment that featured some 150 screenings of Disney films including The Jungle Book, Star Wars: The Force Awakens, and Big Hero 6 in 400-seat theaters. The results were dense with data: some 3,179 people generated 16 million observations that researchers could further study.

Courtesy of Disney Research

Courtesy of Disney Research

The study shows human nature is highly predictable as Disney researchers use factorized variational autoencoders (FVAEs) to read and analyze the individual viewer’s facial expressions. The result? After just a few minutes of observation, the FVAEs adept at predicting individual viewer’s expressions for the remainder of the film.

“The FVAEs were able to learn concepts such as smiling and laughing on their own,” said Zhiwei Deng, lab associate at Disney Research, told Phys.org. “What’s more, they were able to show how these facial expressions correlated with humorous scenes.”

The neural network was able to synthesize vast stores of data that the researchers would never be able to process in a timely manner without losing the important details, thereby providing the team with higher quality results than traditional test screenings.

Courtesy of Disney Research

Courtesy of Disney Research

By using artificial intelligence, filmmakers can re-cut movies to maximize viewers’ pleasure in a way they could have never done before, using the feedback they receive in order to serve, and potentially manipulate, human nature. The more acclimate technology becomes to the pre-verbal emotional space, the more power it gains.

Ultimately, such a tool is profitable in the hands of those who control it, allowing it to create content that reduces critical thinking by honing in on a desire for entertainment to wash away the complex and unnerving truths of reality.


Miss Rosen is a journalist covering art, photography, culture, and books. Her byline has appeared in L’Uomo Vogue, Vogue Online, The Undefeated, Dazed Digital, Aperture Online, and Feature Shoot. Follow her on Twitter @Miss_Rosen.