Extraction of Semantic Keyframes Based on Visual Attention and Affective Models | IEEE Conference Publication | IEEE Xplore