A Survey of Image Emotional Semantic Research
Weining Wang · Journal of Circuits and Systems · 2003
Human perception and understanding of image are subjective. The motivation of image emotional semantic research is to endow the computer the ability for understanding image in emotional way as human being does. The results are to realize image subjective emotional semantic description, these results can be used in fields such as image emotional evaluation, image retrieval, design support, etc. It is a promising and challenging direction for research. In this paper, status quo of image emotional semantic research is reviewed. Typical system architectures are given, key techniques of the system are analyzed and discussed from three aspects: approaches to image feature extraction, schemes of human emotion description, approaches to build the mapping between image features and emotional space.