Discovering Harmony: A Hierarchical Colour Harmony Model for Aesthetics Assessment

Abstract

Color harmony is an important factor for image aesthetics assessment. Although plenty of color harmony theories are proposed by artists and scientists, there is little firm consensus and ambiguous definition amongst them, or even contradictory between them, which causes the existing theories infeasible for image aesthetics assessment. In order to overcome the problem of conventional color harmony theories, in this paper, we propose a hierarchical unsupervised learning approach to learn the compatible color combinations from large dataset. By using this generative color harmony model, we attempt to uncover the underlying principles that generate pleasing color combinations based on natural images. The main advantage of our method is that no prior empirical knowledge of image aesthetics, color harmony or arts is needed to complete the task of color harmony assessment. The experimental results on the public dataset show that our method outperforms the conventional rule based image aesthetics assessment approach.