基于AIGC技术的民族服饰设计研究——以畲族为例  

Research on ethnic costume design based on AIGC technology:Taking the She ethnic group as an example

作　　者：吴海鸣 陈敬玉[1,2] WU Haiming;CHEN Jingyu(School of Fashion Design&Engineering,Zhejiang Sci-Tech University,Hangzhou 311103,China;Zhejiang Sci-Tech University Digital Intelligence Style and Creative Design Research Center,Key Research Center of Philosophy and Social Sciences,Hangzhou 310018,China)

机构地区：[1]浙江理工大学服装学院,杭州311103 [2]浙江省哲学社会科学重点培育研究基地,浙江理工大学数智风格与创意设计研究中心,杭州310018

出　　处：《丝绸》2025年第1期20-29,共10页Journal of Silk

基　　金：浙江省哲学社会科学年度课题项目(22NDJC077YB);浙江省大学生科技创新活动计划(新苗人才计划)项目(2024R406B073)。

摘　　要：民族服饰的当代创新需要在创作过程中寻求民族传统与现代审美的最佳平衡点,生成式人工智能(AIGC)技术的出现为民族服饰的当代设计应用提供了新的路径和方法。文章通过分析目前人工智能技术在民族服饰生成过程中遇到的问题,提出基于专属资源库模型训练的方法并以畲族服饰为例进行实验。实验表明,通过对畲族资源库中的服饰样本进行品类归纳和图像标注进行专属模型的训练,可以使被训练的模型理解、学习到资源库样本中畲族服饰的特征,进而使生成的内容具有畲族服饰风格的图像。通过这一实验,展示了人工智能技术给民族服饰创新设计带来的全新思路和方法,旨在建立一条民族服饰设计与AIGC技术相结合的创新实践路径,通过AIGC技术能促进民族服饰设计的创新性发展和创造性转化。The inheritance and innovation of ethnic clothing are critical aspects of preserving and developing cultural heritage.As modern design demands evolve,the integration of traditional ethnic clothing elements into contemporary fashion faces increasing challenges.Designers must navigate the complex task of balancing the preservation of cultural essence with modern aesthetic integration.The rise of artificial intelligence-generated content(AIGC)technology offers promising solutions for the innovative design and application of ethnic clothing elements.This study focuses on She ethnic group clothing and explores the application of digital generation techniques and innovative design methods through the training of Low-Rank Adaptation(LoRA)and Checkpoints(CKPT)models.A key aspect of this study is the development of a custom model based on a specialized She ethnic group clothing resource database.The clothing samples within this database were systematically categorized and annotated to facilitate the training of the LoRA and CKPT models.These models were specifically trained to recognize,learn,and generate images that accurately depict the distinctive characteristics of She ethnic group clothing.During the training process,deep learning algorithms were employed to optimize the generated images,ensuring high levels of detail fidelity and consistency in style.To evaluate the effectiveness of these models,comparative analyses were conducted to assess the accuracy and alignment of the generated images with traditional She ethnic group clothing designs.The study introduced AIGC technology into the digital translation and innovative design of She ethnic group clothing by leveraging a specialized resource database for model training.This approach ensured the precise capture and creative transformation of traditional She ethnic group clothing elements.The in-depth analysis of the database samples enabled the high-precision reproduction of both style and detail in the generated images,offering a feasible technological pathway for t

关 键 词：AIGC 民族服饰 辅助设计 畲族 Stable Diffusion Low-Rank Adaptation CHECKPOINTS 

分 类 号：TS941.26[轻工技术与工程—服装设计与工程]