Fig. 8From: COGNIMUSE: a multimodal video database annotated with saliency, events, semantics and emotion with application to summarizationSample frames for a low arousal, b high arousal, c very negative valence, and d very positive valenceBack to article page