Estimating human response to taste using EEG.

In order to implement affective computing, there have been several studies to elicit human emotion using audio and video stimuli or by recalling previous events. Taste-elicited emotion has also been investigated using food to induce different levels of pleasure. This is monitored using a range of methods, from questionnaire feedback to electrophysiological responses of autonomic nervous system (ANS) and central nervous system (CNS). In this work, we establish that emotions elicited by taste can be monitored using electroencephalogram (EEG), and, for rigour, compare the response to a taste stimulus against the response to the recall of the same taste. The character of emotions were assessed using a subjective measurement, the hedonic score, which describes the pleasant or unpleasant moods of subjects in response to each taste. The classification performance of EEG responses shows excellent separability between the different emotions induced by different tastes. In addition, it is shown that emotion elicited by taste recall is stronger than the stimulus-elicited emotion.