Simply put,
AI for Games and Animation: A Cognitive
Modeling Approach shows how video
game characters can
be taught to "think" for themselves. This groundbreaking
title points the way for the future of intelligent gaming.
While this book features formal mathematical proofs of its
arguments, it also provides a worthwhile overview of how
today's video game characters are currently programmed, as
well as future possibilities for making them "smarter" (and
more realistic) through techniques borrowed from AI and
robotics. At this level, the text shows off the possibility
of adding
cognitive models (expressed in the
author's own Cognitive Modeling Language, CML) to the games
of the near future.
While today's games feature behavior models and
incorporate the rules of physics to create realistic
scenes, this book makes a strong case that by adding
intelligence to characters, game realism can be enhanced
further while actually simplifying development. Though this
book is certainly worthy as an academic treatise on the
subject, the author keeps an eye on real-world programming
issues. For instance, he suggests that adding AI to game
characters will coexist with current programming
techniques. He is also careful to pay attention to the
limits of CPU and graphics processing power when suggesting
new algorithms.
The most abstract sections here show off the
mathematical techniques (and modeling language) for adding
knowledge representation, sensing, and
learning to characters. Later sections offering two
simulated worlds (featuring T-Rexes and raptors, for
example) show off the author's argument to good
advantage.
Though it requires a background in math to appreciate
fully, AI for Games and Animation can be read
profitably by anyone interested in the future of gaming. It
is only a matter of time before games adopt some of the
techniques presented here, and serious game developers will
get a head start with this intriguing and groundbreaking
book. --Richard Dragan
Topics covered: Cognitive modeling for games and
simulations, geometric and behavior models, domain
knowledge, sensing, interval arithmetic basics, creating
nondeterministic behavior, Cognitive Modeling Language
(CML), and sample model worlds.
Introduces a new approach to creating autonomous
computer-generated characters for games and animation.
Using this technique, an animator or programmer need only
specify a behavior outline, and, through reasoning, the
character will automatically work out a detailed sequence
of actions. Individual chapters provide concrete examples
of advanced character animation, automated cinematography,
and real-time computer game API. Source code, images,
animations, and other resources are available at a web
site.