Perceiving emotions through psychophysiological signals

Emotions influence our cognitive functioning heavily. Therefore, it is interesting to develop measurement techniques that can record experienced emotions. Moreover, to improve user system interaction, computers need to recognize and respond properly to their user's emotional state. This would enable affective computing, which relates to, arises from, or deliberately influences emotion. A range of experiments will be discussed in which a range of psychophysiological measures are applied to penetrate human emotion space. Hereby, we distinguish three facets: the obtrusiveness and noise sensitivity of the measures and the ecological validity of the research. Several statistical parameters were derived from physiological measurements of three electromyography signals: frontalis (EMG1), corrugator supercilii (EMG2), and zygomaticus major (EMG3). In one experiment, 24 participants were asked to watch film scenes of 120 seconds, which they rated afterward. These ratings enabled us to distinguish four categories of emotions: negative, positive, mixed, and neutral. Using the EMG2 and EMG3, discrimination between the four emotion categories was possible. In two other experiments, the 26 participants were asked to read out a story and to relive a recent anxious experience and speak about it. The latter enabled us to determine the amount of experienced arousal. In addition to the three experiments, experiences with galvanic skin conductance and heart rate variability will be discussed. In all instances, real time processing of the signals proved to be possible. This enables tailored user system interaction, facilitated by an emotional awareness of systems. Such systems could, for example, be applied to increase the immersion of participants in games, in ambient intelligence settings, incorporating a Personalized Empathic Computing (PEC), or in telepsychiatry settings. Such systems would introduce a new era in user system interaction.