Into the Uncanny Valley: Bridging the Gap Between Cartoon Avatars and Digital Likeness

Courtesy of Marvel Studios

Courtesy of Marvel Studios

Imagine a scenario where a visual effects product is far too realistic for the human eye and mind to accept. The images look real, true to human form, and yet no matter how similar the resemblance, they’re impossible to believe. The eye recognizes the visual as lifelike, yet the mind refutes the image and rejects an emotional connection.

There’s a narrow intersection, a place where technology and aesthetics meet, that creates this barrier within the brain. It’s a concept known as the “uncanny valley,” and this principle has become one of the last major hurdles in the development of lifelike digital avatars.

Uncanny_valley.png

“As human beings, we are wired innately to understand the human face and all of its subtle expressions,” explains Possible Reality CCO Michael Tigar. “So when we see a digital representation of the human face -- in cartoon form, or all the way to photorealism and beyond -- which goes in to the uncanny valley, we have a built in acceptance of what it is we’re seeing. But when what we’re seeing crosses a certain threshold that we know both consciously and subconsciously when we cross into the uncanny valley, our mind dismisses what we’re seeing. We disengage and we don’t buy into it.”

Sounding more like a philosopher, or psychologist, than a visual effects artist, Tigar has been working on bridging the gap across the uncanny valley since his early days in product prototyping.

Working alongside the acclaimed Dr. Paul Eckman, who created the Facial Action Coding System (FACS) and was also an advisor to James Cameron on the film “Avatar,” Tigar developed a medical application that leaned heavily on the finer points of empathic communication and how the uncanny valley affects human connection and response.

Facial Action Coding System Digital Avatars

“The uncanny valley offers a very realistic barometer of where we need to be, and it gives us a better understanding of why the industry came out with the cartoon representations of ourselves first,” offers Tigar. “If we follow the logic of the uncanny valley, we’re more likely to accept a cartoon version of ourselves than something that is so realistic. There is a curve where we accept it and where we don’t accept the digital likeness of the human form.”

Tigar, who was instrumental in the early advent of talking animals in motion pictures during his years at Rhythm & Hues Studios, is quick to reference films like “The Polar Express” and “Final Fantasy” as examples of character design falling deep into the uncanny valley. However sophisticated the digital likeness and human representations in these films, the public rejected any emotional connection to the characters. And as recently as last week, Thanos, the lead villain in “Avengers: Infinity War,” has also become part of the rhetoric of the uncanny valley, as many question whether the character design was too lifelike for its own good.

Conversely, Tigar also points to Disney Pixar and films like “Toy Story” and “The Incredibles,” explaining that “it’s very easy for us as human beings to find empathy in the portrayal of these CG characters because they’re rendered and perform in a way that doesn’t challenge our subconscious as being realistic or hyper realistic. So we buy into it.”

We have a built in acceptance of what it is we’re seeing. But when what we’re seeing crosses a certain threshold that we know both consciously and subconsciously when we cross into the uncanny valley, our mind dismisses what we’re seeing.
— Michael Tigar, CCO Possible Reality

Navigating the nuances of the uncanny valley, Tigar is now applying this concept to 3-D avatar creation for Possible Reality’s INTARACT application, calculating a specific digital likeness that will satisfy users, creating an emotional connection while suspending disbelief.

Developing an avatar creation system that requires users to upload a single two-dimensional photo, Tigar has successfully designed a digital model that captures true digital likeness, but he is now in the process of “taking the edge off complete realism” to satisfy the undulating curve of the uncanny valley and focusing on the next hurdle in the pursuit of lifelike 3-D avatars: hair.

Stay tuned for more from blogs featuring Possible Reality CCO Michael Tigar as we will dive deeper into the “uncanny valley” and address the creation of lifelike hair for avatars in the coming week.


Have an opinion? Interested in alpha testing our new AR and avatar products? We’d love to hear from you and share our exciting new products. For inquiries, please contact dshapiro@possiblereality.co