I've posted my updated paper & slides on www.jorkin.com. Slides include notes that cover most of what I said, and come in a zip file with all of the videos.
I'm excited to see that UK motion synthesis provider NaturalMotion is finally moving their technology into the real-time arena, which I believe is what they always intended to provide.
CEO Torsten Reil did his Master's thesis on this stuff at the University of Sussex's COGS department (now Informatics) nigh-on ten years ago, and I was keeping an eye on his work in Oxford when I took the same course there a couple of years later. It always seemed a great pity that their first product was an offline tool for synthesising canned animations, when the whole point of motion synthesis is that you don't need to be limited to canned animations! :) Thankfully, the 360 and PS3's brute strength makes real-time synthesis a distinct possibility at last.
I wonder what Ken Perlin will make of Euphoria? His work uses hand-crafted stochastic models driven by authorial goals for an 'illusion of life', whereas Torsten's work uses models inspired by evolutionary robotics and related biological research on the dynamical systems which underlie 'real life' motion.
Just in case anyone missed it, Craig Reynolds posted a comment in a previous thread that I thought deserved mention, as it announces a "PSCrowd" system that will be introduced at GDC.
I hate to post anything that might distract us from the excellent thread Jeff started below, but if anyone reading this is planning to attend GDC this year and/or is interested in crowd simulations, click the link below for more info.
(From Craig Reynolds)
Hi! I can't answer any of Adam's questions about Dead Rising, but I did want to chime in belatedly about next gen crowd densities, as I kept intending to do since Greg earlier post. It is hard to do apples to apples comparisons of crowd simulations, as Adam suggests there is such a range of complexity for the crowd members. How smart are the agents? How detailed are their geometrical representations? What kind of animation is used?
Over the last year I have been looking into high performance crowd/flock simulation on the PS3's Cell architecture. (I work at SCE's US R&D group.) Since I wanted to push on the crowd size parameter, I use a very simple visual model for the individuals in my simulated crowds. Their brain are just some steering behaviors. Keeping that in mind: I am able to run a "boid" simulation of a school with 7000 fish animated at 60fps. My preliminary 2d crowd model runs 10,000 individuals at 60fps. In both cases the behavioral update rate is 1/8 of the frame rate. The fish "bodies" are about 30 polys with some displacement animation in a vertex program.
The general idea is to update the simulation in parallel (on 6 SPEs) using the same spatial hashing used to accelerate proximity queries. For those who want the gory details, I will be talking about this "PSCrowd" system at GDC: http://www.cmpevents.com/GD06/a.asp?option=C&V=11&SessID=2769
Our plan is to show these demos and one with more interesting graphics as part of Phil Harrison's Keynote: http://www.gdconf.com/conference/keynotes.htm#phil
PaulT’s post about Bruce Blumberg’s article got buried in other posts before getting feedback, so I wanted to highlight it again. I think Bruce’s article is timely and brings up good points that merit discussion.
http://www.gamasutra.com/features/20060216/blumberg_01.shtml
If we look at the PC games that have dominated the sales for the past couple years, by far the biggest sellers are World of Warcraft and the Sims. These games couldn’t be more different in regards to AI. WoW has very little AI, and the Sims is nearly all AI, but the AI is quite different than what most of us normally focus on.
Most of us (on this blog) focus on creating believably human behavior. We want AI that can navigate, take cover, flank, and use vehicles as well as the human players. With the rising popularity of multiplayer games, I sometimes wonder what the future holds for us, when there are ample numbers of humans on-line to play with.
What is refreshing about Bruce’s article (and the research of his Synthetic Characters group that it’s based on) is that it offers game AI developers the possibility to create characters that are more alive and believable than the human-controlled characters! Due to interface limitations, it is unlikely we will ever see human-controlled characters who’s eyes widen when they’re surprised, or who’s tail stiffens before a pounce. Human-controlled characters simply do not live in the virtual world the way AI characters do.
So I’m curious what others think about the future of AI in an increasingly multiplayer-dominated world. Will AI be relegated to the roles that humans don’t want to play (e.g. shopkeepers) and “ambient life” (e.g. rats, rabbits, roaches, and crows), or can characters with an inner-life (as Bruce puts it) complement the human characters such that there will always be an equal demand for both?