AI-Assisted Inheritance of Qinghua Porcelain Cultural Genes and Sustainable Design Using Low-Rank Adaptation and Stable Diffusion

作者：Qian Bao, J. Zhao, Ziqi Liu, Na Liang · 发表于：Electronics · 年份：2025 · DOI：10.3390/electronics14040725 · 被引用次数：32 · 研究领域：Color perception and design、Aesthetic Perception and Analysis

Blue-and-white porcelain, as a representative of traditional Chinese craftsmanship, embodies rich cultural genes and possesses significant research value. Against the backdrop of the generative AI era, this study aims to optimize the creative processes of blue-and-white porcelain to enhance the efficiency and accuracy of complex artistic innovations. Traditional methods of crafting blue-and-white porcelain encounter challenges in accurately and efficiently constructing intricate patterns. This research employs grounded theory in conjunction with the KANO-AHP hybrid model to classify and quantify the core esthetic features of blue-and-white porcelain, thereby establishing a multidimensional esthetic feature library of its patterns. Subsequently, leveraging the Stable Diffusion platform and utilizing Low-Rank Adaptation (LoRA) technology, a generative artificial intelligence (AIGC)-assisted workflow was proposed, capable of accurately restoring and innovating blue-and-white porcelain patterns. This workflow enhances the efficiency and precision of pattern innovation while maintaining consistency with the original artistic style. Finally, by integrating principles of sustainable design, this study explores new pathways for digital innovation in blue-and-white porcelain design, offering viable solutions for the contemporary reinvention of traditional crafts. The results indicate that AIGC technology effectively facilitates the integration of traditional and modern design approach...