ABSTRACT
Visual aesthetic quality is a measure of visually perceived beauty. Judgment of the visual aesthetic quality of images is highly subjective, involving sentiments and personal taste [1]. However, some images are often believed, by consensus, to be visually more appealing than others. This serves as one of the principles in the emerging research area of computational
aesthetics. Computational aesthetics is concerned with exploring computational techniques to predict emotional response to a visual stimulus, and developing methods to create and enhance pleasing impressions [2], [3]. This chapter focuses on exploring computational solutions to automatically infer the aesthetic quality of images. The greatest challenge in this research lies in the gap between low-level image properties and the high-level human perception of aesthetics.
