Bayesian Programming: behaviours for synthetic video games characters

bot (AVI, 6 Mo)

This movie shows an application of Bayesian Programming to behaviours
for synthetic video games characters. We address the problem of
real-time reactive selection of elementary behaviours for an agent
playing a first person shooter game (namely Unreal Tournament). We
show how Bayesian Programming can lead to condensed and easier
formalisation of finite state machine-like behaviour selection, and
lend itself to learning by imitation, in a fully transparent way for
the player.

More precisely, the movie shows a bot (first of all in third-person
view, then in first-person view) fighting autonomously, based on a
behaviour showed by a human. You can see it attacking, retreating,
fetching a health pack... The human showed the behaviour by playing
normally the game for 5 to 10 minutes. The bot behaviour consists in
an FSM-like action selection, implemented in a bayesian fashion. Learning
this model implies recognising elementary behaviours in the human's
actions.