A graphic design proposal for interpreting results in affective experimentations

Recent advances in video technology and deep learning methodologies have fostered the research in emotion recognition, the creation of dedicated databases and the set-up of ad hoc experiments. The adoption of different techniques to assess the emotion, from self-reports to facial expression recognition algorithms, has fuelled research on representing and comparing achieved results. Inspired by Russell’s circumplex model of affect, this study proposes a novel graphic design based on balloon plot concept to incorporate and compare results from two different assessment techniques. As a case study, we apply this design to the creation of a spontaneous facial expression database, where self-reports are compared to expected results. A user study has been conducted, and a validation was carried out, revealing high perceived usability among participants. The result is an immediate graphical representation with qualitative and quantitative significance that could be customised depending on the experiment, thereby providing a compact alternative to traditional representations.