Image Style Transfer Based on Deep Learning Network
Jingyang Qiao, Haixin Wang, Zejun Zhao · 2021 2nd International Seminar on Artificial Intelligence, Networking and Information Technology (AINIT) · 2021
Image style transfer is an intersection field of art and computer science, which uses deep learning tools to achieve style transfer between images. The objective of this review is to compare two popular image style transfer algorithms. Specifically, the paper first walks through the history and basic principles of these two algorithms and then discusses several existing drawbacks. After that, the paper introduces the person who proposes the solution to the drawbacks mentioning above. Finally, the paper compares the advantages and disadvantages of each algorithm and discusses the possible future improvements. The main contribution of this review paper is that it provides readers a clear walk-through of today’s popular algorithms of image style transfer as well as explores the potential combined fields of art and computer science by discussing the possible improvements of the existing popular algorithms.