Paint-your-mind II: stylistic image creation by concept

Abstract

The rise of digitalisation in every sector brought on by the Fourth Industrial Revolution has prompted development in the sector of Art in Artificial Intelligence. Much research and progress has been made in the computer vision and image processing field. There is much attention in authentic image synthesis and style transfer using neural networks in recent years. These works have transformed the way art is being created and consumed by the masses. This project explores existing networks and solutions in the field of art (visual art) synthesis. While much research has been made to generate authentic art, few applications are made available to end-users. As such, this project aims to enhance existing applications and provide a meaningful platform which maximises the developments in artificially intelligent art. Hence, this project utilises two methods optimised from current state-of-the-art image translation methods to develop an application that is accessible and simple to use by the masses. The implementation consists of careful research about the image translation networks available and integrating relevant networks to a usable interface. Overall, this project will simplify the process of art creation without impeding on the value of art.