EEG-based Inter-Subject Correlation Schemes in a Stimuli-Shared Framework: Interplay with Valence and Arousal

Affective computing is confronted to high inter-subject variability, in both emotional and physiological responses to a given stimulus. In a stimuli-shared framework, that is to say for different subjects who watch the same stimuli, Inter-Subject Correlation (ISC) measured from Electroencephalographic (EEG) recordings characterize the correlations between the respective signals at the different EEG channels. In order to investigate the interplay between ISC and emotion, we propose to study the effect of valence and arousal on the ISC score. To this end, we exploited various computational schemes corresponding to different subsets of the dataset: all the data, stimulus-wise, subject pairwise, and both stimulus-wise and subject pairwise. We thus applied these schemes to the HCI MAHNOB and DEAP databases. Our results suggest that the ISC score decreases with valence and increases with arousal, as already shown by previous results on functional MRI.