Research on the Digital and Intellectual Development of Tie-Dye Patterns Based on Deep Learning

Abstract

Tie-dye art, as an important expression of traditional Chinese culture, carries unique craft wisdom and aesthetic value. However, traditional tie-dye relies on experience-based creation and involves complex processes, making it difficult to effectively meet the demands of the modern design industry for efficient production and diverse expression. Against the backdrop of the rapid development of Artificial Intelligence Generated Content (AIGC) technology, deep learning offers new possibilities for the digital innovation of traditional patterns. This paper focuses on the application of deep learning in the digital and intelligent development of tie-dye patterns. It provides an in-depth review of innovative approaches utilizing technologies such as Generative Adversarial Networks (GANs) and diffusion models in pattern generation, texture expression, and color control. Furthermore, it analyzes their specific applications in the intelligentization of craft processes, including the recognition of tying structures, prediction of diffusion effects, and quality inspection of finished products. The study reveals that AIGC technologies, represented by deep learning, are progressively advancing tie-dye art from an experience-driven to a data-driven paradigm. They provide effective technical support for the innovative expression and digital dissemination of traditional crafts, constituting a crucial pathway for the modernization and transformation of tie-dye art.