Oculus Chief Scientist Michael Abrash says the industry needs to improve the way individuals are represented – in all their uniqueness – within a VR framework.
“Other people are what we are most highly tuned to, because they are what we care about most – and for that same reason, representing them believably is one of the greatest challenges,” he explained in a statement quoted by UploadVR.
Image Credit: UploadVR
“In the long run, once virtual humans are as individually quirky and recognizable as real humans, VR will be the most social experience ever, allowing people anywhere on the planet to share virtually any imaginable experience.”
According to Abrash, virtual technology has already started to revolutionize everything from education to telepresence to architecture to gaming, and well beyond.
“This will continue for decades, in ways we haven’t even dreamed of yet, with ever greater impact on how we work, play, and interact with one another,” he added.
Image Credit: Oculus
Commenting on the above, Rambus Fellow Dr. David Stork tells us that viewers have historically been quite comfortable watching virtual toys, figures, or avatars on-screen so long as they don’t strive to be precisely humanoid.
“If, however, the figures are nearly—but not quite accurately—humanoid, viewers find such figures annoying or even repulsive,” he explained. “This phenomenon has been called uncanny valley: the deep drop in the viewers’ acceptance of figures that are visually humanoid, but not sufficiently similar to an actual human.”
The popular success of Pixar’s first feature film, Toy Story (1995), says Stork, was due in part to the fact that the central characters were toys, and that the “human” characters, such as Andy and his sister, were shown only rarely or, in the case of the vicious neighbor Sid, meant to be repulsive in their own right.
Image Credit: Pixar
“The magnitude of this repulsive effect—the depth of the uncanny valley—is far larger for animated or moving characters than it is for static characters,” he continued. “This is in part why people are quite comfortable viewing traditional (static) oil portraits, even if the portrait isn’t objectively similar to the real human subject. The uncanny valley for moving humanoid characters presents a deep challenge for makers of virtual reality hardware and especially software.”
Viewers, Stork emphasizes, are exquisitely sensitive to the subtlest variations in expressions—even subliminal “micro-expressions.”
“After all, we’ve evolved over millions of years to infer the intentions of potential adversaries and mates based on their facial expression. Traversing the uncanny valley using low-cost computational hardware (instead of the supercomputers of animation studios) and real-time algorithms (rather than the long production periods of feature films) will make virtual reality immersive displays compelling, particularly for gaming. There’s clearly large customer demand for such systems,” he concluded.