Journal of Engineering, Project, and Production Management, 2026, 16(7), 2025-308
Artistic Style Transfer Based on Generative Adversarial Networks and Its Application in Film and Animation Production
Lecturer, School of Film, Jilin Animation Institute, Changchun 130000, China, E-mail: wangshifeng2025@163.com
Production Management
Received December 7, 2025; revised January 27, 2026; accepted July 2, 2026
Available online July 12, 2026
Abstract: Artistic style transfer technology plays an important role in film and animation production. High-quality artistic style transfer can significantly enhance the production value of animation. However, traditional methods for artistic style transfer face limitations, including low efficiency and poor results. Therefore, this study proposes a model for artistic style transfer in film and animation production based on Deep Convolutional Generative Adversarial Networks (DCGAN). The model utilizes the Feature Pyramid Network (FPN) with a skip connection structure and introduces sub-pixel convolution blocks and Batch Normalization to further improve the generator architecture. To further reduce interference from abnormal data, an improved multi-window spectral estimation and spectral subtraction method is introduced to enhance data representation, thereby enabling high-quality film and animation production. Experimental results show that the proposed algorithm achieves high adaptability and strong fitting performance, with content and style losses of 0.09 and 0.12, respectively and outperforms the comparison algorithms. The constructed model also demonstrates strong performance in artistic style classification and transfer. In two datasets, the matching degree of style features was 89.14% and 91.20%, respectively. The proposed model can maintain high-quality style transfer under various forms of noise, and the ablation study demonstrates its validity. These results indicate that the artistic style transfer model can enhance the efficiency and accuracy of style recognition while improving animation quality. This study contributes to the precise construction of artistic style transfer models for film and animation production and promotes the development of the animation industry.
Keywords: Deep convolutional generative adversarial networks; generator; artistic style transfer; animation production; multi-window spectral subtraction.
Copyright © Journal of Engineering, Project, and Production Management (EPPM-Journal).
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Requests for reprints and permissions at eppm.journal@gmail.com.
Citation: Wang, S. (2026). Artistic Style Transfer Based on Generative Adversarial Networks and Its Application in Film and Animation Production. Journal of Engineering, Project, and Production Management, 16(7), 2025-308.
