Affect Recognition during Active Game Playing based on Posture Skeleton Data.

摘要: The affective state of a player during game playing has significant effect on the player’s motivation and engagement. Recognising emotions games can help designers improve user experience by providing sophisticated behaviours to characters system itself. This paper presents work-in-progress towards novel recognition using posture skeleton data as input from non-intrusive interfaces. A database samples non-acted was captured active Microsoft Kinect’s sensors. Four observers were asked annotate selected postures with an emotion label given set. Based Cohen’s kappa, agreement level above or equal ‘good’ overall levels that outperform existing benchmarks. used in series experiments for training recognising emotions. results indicate compiled annotated labels performs considerably chance offers interesting research questions improvements future directions area.