Journal of Engineering, Project, and Production Management, 2026, 16(7), 2026-0061
Face Replacement Method for Film and Television Drama Videos Based on Generative Adversarial Network and Task Decomposition Strategy
1 Associate
Professor, Hunan University of Media and Communications, China
2 Department Head, Indoor Choir in
the School of Art, Shijiazhuang Institute of Technology, China, E-mail:
zhouulii@outlook.com (corresponding author).
3 Director, School of Art,
Shijiazhuang Institute of Technology, China
Production Management
Received May 29, 2026; revised July 14, 2026; accepted July 23, 2026
Available online August 12, 2026
Abstract: The current face replacement technology faces problems such as insufficient detection accuracy and poor image quality when processing high-resolution videos. A face replacement method for film and television videos, grounded in improved multi-task Convolutional Neural Networks (CNN) and Generative Adversarial Networks (GANs), is proposed to enhance face detection accuracy and the naturalness and realism of replacement images. An improved Multi-Task Convolutional Neural Network (MTCNN) algorithm that incorporates depthwise separable convolution and a feature pyramid structure is studied. A generator-discriminator structure based on an improved generative adversarial network is designed, with residual blocks introduced to enhance performance. The research findings indicate that the raised generative adversarial network achieves a Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) of 0.10, a peak signal-to-noise ratio close to 25.0dB, a structural similarity index close to 1.0, and an emotion retention rate of 85%. Meanwhile, its computing resource consumption is 19%, and its robustness is 90%, indicating excellent performance. The proposed method for facial replacement in film and television dramas can raise the efficiency and quality of facial replacement, providing technical support for post-production and video content creation, as well as new ideas and methods for related technology research.
Keywords: GAN; task decomposition strategy; face replacement; film and television drama videos; facial detection.
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: Shan, L., Zhou, L., and Du, Y. (2026). Face Replacement Method for Film and Television Drama Videos Based on Generative Adversarial Network and Task Decomposition Strategy. Journal of Engineering, Project, and Production Management, 16(7), 2026-0061.
