Painting-Emotion Matching Technology Learning System through Repetition

Abstract

People’s interest in paintings has increased as artists have easier access to an audience. However, at times, laypersons may not understand the significance of a painting. With the development of computer science, it has become possible to analyze paintings using machines, but some limitations remain. In this paper, we present a learning tool to help analyze the sensitivity of a given painting. To this end, the proposed system provides users with the ability to predict the emotions expressed by a painting through repeated learning of a matched painting. Using this learning tool, users can improve their ability to understand paintings.

Emotion-Inspired Painterly Rendering

“…Lee et al [2] extracted three representative colors of painting and defined emotional values of painting by matching emotional adjectives with the three colors defined by Kobayashi. And Lee's work [3] presented a learning tool that increases the ability to analyze the sensibility of paintings based on CNN. Through this system, non-experts can improve their understanding of painting's emotions by repeated learning to choose the picture that is closest to a given picture and emotion.…”