Scholarlink

<p>As an integral part of Chinese culture, traditional cheongsam pattern faces challenges in inheritance and innovation. This study uses deep learning technology to carry out innovative design of traditional cheongsam patterns, giving modern fashion sense while retaining cultural essence. By collecting and sorting 1000 traditional cheongsam pattern images for data cleaning and prepossessing, a constitutional neural network and generative adversarial network were used for feature extraction and pattern generation, and multiple optimization was carried out combined with user feedback. The results show that the innovative design obtains high user satisfaction in the aspects of visual beauty, cultural inheritance and modern innovation. This study provides a new method and path for the modern inheritance and digital preservation of traditional culture, which has important theoretical and practical significance.In a user satisfaction survey conducted with 828 valid responses, more than 70% of respondents expressed satisfaction with the innovative designs. The survey reported an average satisfaction rating of 4.2 out of 5 for visual beauty, 3.9 for cultural inheritance, and 4.0 for modern innovation. This study provides a new method and path for the modern inheritance and digital preservation of traditional culture, which has important theoretical and practical significance.</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

Innovative design of traditional cheongsam pattern based on deep learning

  • 期刊论文

  • 发表日期：

    2025年8月27日

查看全文
  • Qiong Luo,
  • Yun Pan,
  • Xubing Xu

摘要

As an integral part of Chinese culture, traditional cheongsam pattern faces challenges in inheritance and innovation. This study uses deep learning technology to carry out innovative design of traditional cheongsam patterns, giving modern fashion sense while retaining cultural essence. By collecting and sorting 1000 traditional cheongsam pattern images for data cleaning and prepossessing, a constitutional neural network and generative adversarial network were used for feature extraction and pattern generation, and multiple optimization was carried out combined with user feedback. The results show that the innovative design obtains high user satisfaction in the aspects of visual beauty, cultural inheritance and modern innovation. This study provides a new method and path for the modern inheritance and digital preservation of traditional culture, which has important theoretical and practical significance.In a user satisfaction survey conducted with 828 valid responses, more than 70% of respondents expressed satisfaction with the innovative designs. The survey reported an average satisfaction rating of 4.2 out of 5 for visual beauty, 3.9 for cultural inheritance, and 4.0 for modern innovation. This study provides a new method and path for the modern inheritance and digital preservation of traditional culture, which has important theoretical and practical significance.