by Shangfei Wang, Qiang Ji

Abstract?Video affective content analysis has been an active researcharea in recent decades, since emotion is an important componentin the classification and retrieval of videos. Video affective contentanalysis can be divided into two approaches: direct and implicit.Direct approaches infer the affective content of videos directly fromrelated audiovisual features. Implicit approaches, on the other hand,detect affective content from videos based on an automatic analysisof a user?s spontaneous response while consuming the videos. Thispaper first proposes a general framework for video affective contentanalysis, which includes video content, emotional descriptors, andusers? spontaneous nonverbal responses, as well as the relationshipsbetween the three.