Visual content learning for visualizations memorability classification

Abstract

Images on the Internet and in multimedia systems are rising successively. There are different research works on visual information and automatic analysis of images. Image memorability is a new task in computer vision. Actually, the human brain processes simultaneously millions of images and other information from multiple sources. Among these various images and information some of them are more memorable than others. Studying images memorability is an image processing task, which tends to define and establish the characteristics of memorable images. These characteristics are used to create a representative model for predicting images memorability. In this paper, we present an approach for visualizations memorability analysis and evaluation. We use a set of features that are automatically extracted from visualizations. Then, we employ different classification methods to explore these features and visualizations scores.