Proposal for a saliency map based on the focal points of a painting

We assumed that there was regularity in the eye movements of viewers of a good painting. By constructing a database of gaze-point distributions, we found that the relationship between the author's intention and the gaze points of the viewer can be clarified, and we are creating a gaze-point-distribution database. In making a painting, the painter attempts to guide the viewer's eye movement. We studied the degree of coincidence between gaze points and the saliency map obtained by filtering the original image on the basis of the painting’s focal point. We believe it will be possible to predict the distribution of viewer’s gaze points when looking at a painting with high accuracy using the proposed method.