| Sensing Emotions | |
 |
Sensor technology can obtain a broad range of signals, related to the affective state of people (e.g., an emotion). However, people’s affective states are fuzzy, ill-defined. So, crucial is not only capturing high quality signals but also their mapping to sensible psychological constructs. Moreover, multiple signals are claimed to reveal distinct affective states. Self-evidently, the signals of interest are also influenced by numerous sources of noise (e.g., movement and other signal sources), as is with every real-world application. The focus of the presentation will be on tackling the issues mentioned, when measuring biosignals (i.e., physiological signals) in (real-world) ambulatory settings. Combinations of biosignals and their combination with other signals, in particular speech and visual signals will be explored. Also, the accuracy of human emotion recognition will be compared with machine’s emotion recognition. Several examples of affective signal processing technology will be given that emerged from a series of projects. |
 |
Egon van den Broek, University Utrecht |