Fig. 11From: COGNIMUSE: a multimodal video database annotated with saliency, events, semantics and emotion with application to summarizationObjective evaluation comparative results for TMM13 and ICIP15, where saliency classification ROC curves for the different modalities are presentedBack to article page