Article: Optimisation of art image style migration algorithms based on deep neural networks Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.33 pp.18 - 37 Abstract: As deep learning technology has grown quickly, art image style migration, which is an important area of research in computer vision, has become more and more common. The conventional style migration technique exhibits issues such as inconsistent outcomes and diminished computing efficiency. This research presents a deep neural network (DNN)-based optimisation framework for art image style migration techniques, with the objective of optimising the style migration effect while minimising computational overhead and improving operational efficiency. The suggested algorithm demonstrates excellent migration effects and high efficiency on two real experimental datasets by optimising the network topology, integrating varied loss functions, upgrading the style expression mechanism, and improving operational efficiency. The experimental results demonstrate that the suggested algorithm surpasses existing mainstream approaches in various respects and possesses significant practical application value. Finally, the study talks about the algorithm's flaws and suggests areas for future research. Inderscience Publishers - linking academia, business and industry through research

Title: Optimisation of art image style migration algorithms based on deep neural networks

Authors: Xiangkun Gao

Addresses: School of Design and Art, Jingdezhen Ceramic University, Jingdezhen 333000, China

Abstract: As deep learning technology has grown quickly, art image style migration, which is an important area of research in computer vision, has become more and more common. The conventional style migration technique exhibits issues such as inconsistent outcomes and diminished computing efficiency. This research presents a deep neural network (DNN)-based optimisation framework for art image style migration techniques, with the objective of optimising the style migration effect while minimising computational overhead and improving operational efficiency. The suggested algorithm demonstrates excellent migration effects and high efficiency on two real experimental datasets by optimising the network topology, integrating varied loss functions, upgrading the style expression mechanism, and improving operational efficiency. The experimental results demonstrate that the suggested algorithm surpasses existing mainstream approaches in various respects and possesses significant practical application value. Finally, the study talks about the algorithm's flaws and suggests areas for future research.

Keywords: art images; deep neural network; DNN; style migration optimisation; algorithm optimisation.

DOI: 10.1504/IJICT.2025.148656

International Journal of Information and Communication Technology, 2025 Vol.26 No.33, pp.18 - 37

Received: 22 Jun 2025
Accepted: 15 Jul 2025
Published online: 17 Sep 2025 *

Title: Optimisation of art image style migration algorithms based on deep neural networks

Keep up-to-date