Hell, as Jean-Paul Sartre rightly noted, is other people. That is emphatically true in real life and apparently also true in virtual reality, where simulating the movement of crowds is a trying task.
If that is your kind of topic, here is an excellent podcast interview with Rutgers’ professor Mubbasir Kapadia about his upcoming book, Simulating Heterogeneous Crowds with Interactive Behaviors. I will spare you the finer technical points because that’s what the podcast is for, but you should definitely give it a listen.
Animated crowds have a long and occasionally ignominious history. In nearly any recent film with large crowd scenes, most of those people are computer generated. Here, for instance, is a good breakdown of a scene from the fabulous El secreto de sus ojos (not to be confused with the English-language remake, The Secret in Their Eyes):
From a certain distance—and in certain uses—CGI crowds work just fine. People in soccer stadia behave in broadly predictable manners, so as long as the camera is not staring straight at a CGI section, there is no real problem with repeating a small patch of human actors or animating characters with a certain amount of noise and variance added to their movements. The same holds true for videogames. Non-player characters with limited roles needn’t be the biggest impediments to a game’s believability. Lifelike behavior is understandable enough to populate a world in this manner.
But in virtual reality, things are a bit more complicated. (Isn’t it always thus?) There are more ways to interact with more characters, and to do so up close. All of this requires more coordination. There are more opportunities to notice unrealities in VR, which helps to explain why more advanced crowd simulations are needed. It’s yet another example of established techniques needing refinement to maintain their verisimilitude in virtual reality.
Header image via Flickr.
