Affective games: a multimodal classification system

Affective gaming is a relatively new field of research that exploits human emotions to influence gameplay for an enhanced player experience. Changes in player’s psychology reflect on their behaviour and physiology, hence recognition of such variation is a core element in affective games. Complementary sources of affect offer more reliable recognition, especially in contexts where one modality is partial or unavailable. As a multimodal recognition system, affect-aware games are subject to the practical difficulties met by traditional trained classifiers. In addition, inherited game-related challenges in terms of data collection and performance arise while attempting to sustain an acceptable level of immersion. Most existing scenarios employ sensors that offer limited freedom of movement resulting in less realistic experiences. Recent advances now offer technology that allows players to communicate more freely and naturally with the game, and furthermore, control it without the use of input devices. However, the affective game industry is still in its infancy and definitely needs to catch up with the current life-like level of adaptation provided by graphics and animation.