A Behavioral Study of Improved Style Transfer Network with Various Textural Styles Applied to a Female Human Face

Abstract

Art is a required class in many countries. It helps people learn and construct ideas more visually. Recognizing patterns and textures is a skill that challenges even many art teachers when instructing their students. Based on a neural network study called Improving the Neural Algorithm of Artistic Style, the idea of artistic transfer portrays the future that computers can be used to ameliorate the hardship some teachers find. In this research, we specifically use four patterns that have distinct styles and illustrate some basic rules on how to apply these style images. We also examine the styling behavior when using grayscale content images.