*2.6. Implicit Emotions by Automated Facial Expression Response Measurement*

The facial expressions of 30 among 47 panellists were evaluated using Affdex (Affectiva Inc., Waltham, MA, USA) based on the facial inputs. The automated facial coding engine (AFFDEX) was integrated with iMotions Facial Expression Analysis Module (iMotions, Inc., Boston, MA, USA) for decoding the facial emotions using a group of action units (Table 1). The iMotions Facial Expression Analysis Module detects and extracts seven core emotions (joy, anger, fear, disgust, contempt, sadness, and surprise) (shown in Figure 1) and 20 facial expression measures (action units). The action units describe the movements of facial muscles. Emotions are displayed by the movements of a certain number of combined facial muscles. iMotions module also provides timelines annotations, and scores of engagement and valence to provide insight into the facial emotions (https://imotions.com/). The intensity for emotional expression varies from 0 (no expression) to 100 (expression present). The facial expression data collected was from the first three seconds after the participants put the energy drink in their mouths. This is based on previous studies in drinks, in which it was demonstrated that automatic nervous system (AND) responses are expressed immediately after participants are exposed to the stimuli [37,51].


**Table 1.** Action units of the facial muscle movements expressing emotions in Affective® https://imotions.com/blog/facialaction-coding-system/.

**Figure 1.** Different emotions: sadness, happy (joy), surprise, disgust, anger, fear, and contempt elicited by a participant. This figure shows 12 s of recording as a demonstration of the different faces' movements and their relationship with emotions. Actual facial expressions were taken during three seconds after swallowing the sample.
