Automated Emotional Facial Expression Assessment And Emotional Elicitation Through Film Clip Stimuli

Numerous film clip databases are available for eliciting emotional states in humans. Some of the databases have been validated through self-reported questionnaires based on the discrete emotions perspective. In this study we analyzed some of these film clips using a software to assess emotional facial expression in humans. To do so, we selected 12 emotional stimuli (two for each emotion assessed). Other film clips containing basic mathematical operations were used as distractor stimuli. In total, 65 healthy volunteers participated in this study. We performed statistical analyses to compare differences in the discrete emotional intensities of each stimulus and compared these intensities with the distractor stimuli. Although the emotional facial recognition software was able to clearly detect discrete emotions for some stimuli (happiness and anger), some inconsistencies were found between previous self-reported emotional assessments studies and the data obtained with this software. Our results also showed that film clip stimuli present a complex emotional profile, making it difficult to classify them into discrete categories. Software to detect facial emotional expression may therefore be a useful tool for investigating emotions and the emotional profiles of film clip stimuli. However, further studies are needed to corroborate our results.