Gender Effects on an EEG-Based Emotion Level Classification System.

Emotion level classification systems based on features extracted from physiological signals have promising applications in human-computer interfaces. Moreover, there is increasing evidence that points to gender differences in the processing of emotional stimuli. However, such differences are commonly overlooked during the assessment and development of the systems in question. Here, we study gender differences in the performance of an emotion level classification system and its constituting elements, namely features extracted from electroencephalography (EEG) signals, and emotion level ratings in the Arousal/Valence (AV) dimensional space elicited from audiovisual stimuli. Obtained results show differences in the physiological and expressive responses of men and women, and in overall classification performance for the valence dimension.