Physiological Sensing of Emotion

Physiological changes have long been associated with emotion. Although the relative role of cognitive versus physiological processes in emotion has been debated, it is acknowledged that in almost all cases, measurable physiological changes co-occur with emotion, for example: changes in heart rate, sweat level, muscle tension, breathing rate, facial expression and electrical activity in the brain. By sensing these changes we can hope to build computer systems that can automatically recognize emotion by recognizing patterns in these sensor signals that capture physiological responses. This chapter will give a detailed introduction to measuring physiological signals that reflect affect (emotion) with a focus on measuring cardiac activity and skin conductance. The discussion will include why these signals are important for measuring emotional activity, how they are most commonly measured, which features are most often extracted for use in recognition algorithms and the trade-offs between signal quality and wearability and convenience for different sensing systems. Keywords/key phrases: physiological, emotion, heart rate variability, galvanic skin response, signals, sensing