Ambiguity or uncertainty is a pervasive element of many real world decisionmaking processes. Variation in decisions is a norm in this situation when thesame problem is posed to different subjects. Psychological and metaphysicalresearch had proven that decision making by human is subjective. It isinfluenced by many factors such as experience, age, background, etc. Sceneunderstanding is one of the computer vision problems that fall into thiscategory. Conventional methods relax this problem by assuming scene images aremutually exclusive; and therefore, focus on developing different approaches toperform the binary classification tasks. In this paper, we show that sceneimages are non-mutually exclusive, and propose the Fuzzy Qualitative RankClassifier (FQRC) to tackle the aforementioned problems. The proposed FQRCprovides a ranking interpretation instead of binary decision. Evaluations interm of qualitative and quantitative using large numbers and challenging publicscene datasets have shown the effectiveness of our proposed method in modelingthe non-mutually exclusive scene images.
展开▼