Scholarlink

<p>The maintenance of Intangible Cultural Heritage (ICH) clothing remains a significant issue due to the decline of traditional craft skills and the increasing reliance on less sophisticated computational skills. The proposed research presents an Artificial Intelligence (AI)-driven Talent Development System that is based on the Genetic Algorithm tuned Bidirectional Gated Recurrent Units (GA-BiGRU) to preserve and reshape ICH clothing. The system leverages the Cultural Clothing Patterns Dataset, comprising 6740 images of authentic motifs; Miao butterfly embroidery styles, fabric textures, and regional color schemes, organized by motif categories. Gaussian noise reduction–based data preprocessing, and histogram equalization are evened to make images clearer and color even. Histogram of Oriented Gradients (HOG) is employed to extract fine-grained structural and motif-level features from traditional clothing patterns, effectively capturing edge orientation and geometric characteristics. The GA-BiGRU model consists of merging the optimization power of GAs with the temporal power of BiGRU networks to learn the patterning of context and structural designs. Experimental analysis shows that it exhibits good results in terms of motif structural similarity (94%) and Structural Similarity Index Measure (SSIM) of 0.96. Implementing the system is done in Python with the help of deep learning (DL) through the use of Python-based frameworks like TensorFlow, PyTorch, and Open CV to extract features and optimize the system. The suggested Talent Development System is efficient in terms of improving design innovation, advancing AI-based learning, and maintaining cultural sustainability in the design of heritage-based clothes.</p> Graphical abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI and genetic algorithm-based talent development system for intangible cultural heritage clothing preservation and design

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月24日

查看全文
  • Lin Wang,
  • Xue Li

摘要

The maintenance of Intangible Cultural Heritage (ICH) clothing remains a significant issue due to the decline of traditional craft skills and the increasing reliance on less sophisticated computational skills. The proposed research presents an Artificial Intelligence (AI)-driven Talent Development System that is based on the Genetic Algorithm tuned Bidirectional Gated Recurrent Units (GA-BiGRU) to preserve and reshape ICH clothing. The system leverages the Cultural Clothing Patterns Dataset, comprising 6740 images of authentic motifs; Miao butterfly embroidery styles, fabric textures, and regional color schemes, organized by motif categories. Gaussian noise reduction–based data preprocessing, and histogram equalization are evened to make images clearer and color even. Histogram of Oriented Gradients (HOG) is employed to extract fine-grained structural and motif-level features from traditional clothing patterns, effectively capturing edge orientation and geometric characteristics. The GA-BiGRU model consists of merging the optimization power of GAs with the temporal power of BiGRU networks to learn the patterning of context and structural designs. Experimental analysis shows that it exhibits good results in terms of motif structural similarity (94%) and Structural Similarity Index Measure (SSIM) of 0.96. Implementing the system is done in Python with the help of deep learning (DL) through the use of Python-based frameworks like TensorFlow, PyTorch, and Open CV to extract features and optimize the system. The suggested Talent Development System is efficient in terms of improving design innovation, advancing AI-based learning, and maintaining cultural sustainability in the design of heritage-based clothes.

Graphical abstract