Journal of Engineering, Project, and Production Management, 2027, 17(2), 2026-211
Intelligent Decorative Images Design Based on Improved GAN and Stitch Data
Lecturer, School of Digital Media and Art and Design, Jiangsu Vocational Institute of Commerce, Nanjing, 211168, China, E-mail: wangglel@outlook.com
Production Management
Received February 8, 2026; revised April 7, 2026; September 14, 2026; accepted September 24, 2026
Available online October 20, 2026
Abstract: With the growing demand for personalized decoration, traditional manual design methods are inefficient and lack adaptability. This study proposes an intelligent design model for decorative images by integrating stitch data into an improved Generative Adversarial Network (GAN). The model converts stitch data into adjustable parameters that control style and texture, enhancing generation guidance through a dual-discriminator mechanism, stitch-style loss, and a gradient penalty. Convolutional Neural Networks (CNNs) are incorporated for multi-level feature extraction. On the EmbroideryArt-10k and PatternTexture-8k datasets, the model achieved rapid convergence (loss 0 after 500 epochs), high recognition integrity (98.7%), and F1 scores up to 0.92. The model improves design efficiency and offers a new approach to intelligent, personalized decorative image design. In summary, the model not only improves the efficiency of decorative image design but also offers new ideas for the intelligence and personalization of decorative image design, with important theoretical and practical significance.
Keywords: Decorative image design; adversarial network design; parameterized model; convolutional neural networks (CNN); stitch data.
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, L. (2027). Intelligent Decorative Images Design Based on Improved GAN and Stitch Data. Journal of Engineering, Project, and Production Management, 17(2), 2026-211.
