“BioVid Emo DB”: A multimodal database for emotion analyses validated by subjective ratings

摘要

The precondition of productive data mining is having an efficient database to work on. The BioVid Emo DB is a multimodal database recorded for the purpose of analyzing human affective states and data mining related to emotion. Psychophysiological signals such as Skin Conductance Level, Electrocardiogram, Trapezius Electromyogram and also 4 video signals were recorded. 5 discrete emotions (amusement, sadness, anger, disgust and fear) were elicited by 15 standardized film clips. 94 participants watched them, rated them in terms of the experienced emotional level and selected the film clips that evoked the strongest emotion. A preliminary analysis of the subjective ratings made during the experiment is presented. The dataset is available for other researchers.