Does your profile say it all? Using demographics to predict expressive head movement during gameplay

In this work, we explore the relation between expressive head
movement and user pro¯le information in game play settings. Facial ges-
ture analysis cues are statistically correlated with players' demographic
characteristics in two di®erent settings, during game-play and at events
of special interest (when the player loses during game play). Experi-
ments were conducted on the Siren database, which consists of 58 par-
ticipants, playing a modi¯ed version of the Super Mario. Here, as player
demographics are considered the gender and age, while the statistical
importance of certain facial cues (other than typical/universal facial ex-
pressions) was analyzed. The proposed analysis aims at exploring the
option of utilizing demographic characteristics as part of users' pro¯l-
ing scheme and interpreting visual behavior in a manner that takes into
account those features.