文献类型：会议论文

英文题名：Innovative Design of Chinese Traditional Textile Patterns Based on Conditional Generative Adversarial Network

作者：Liu, Miao[1];Zhou, Binbin[1]

机构：[1]East China Univ Sci & Technol, Shanghai 200237, Peoples R China

会议论文集：10th International Conference on Culture and Computing (C and C) Held as Part of the 24th International Conference on Human-Computer Interaction (HCII)

会议日期：JUN 26-JUL 01, 2022

会议地点：ELECTR NETWORK

语种：英文

外文关键词：Chinese traditional textile pattern; Conditional generative adversarial networks; Deep learning; Image classification generation

摘要：With the continuous development of society and economy, Chinese traditional patterns with regional and ethnic characteristics have received more and more attention in international cultural exchanges, and the long history of Chinese traditional patterns has become a valuable resource for design creation. As a valuable cultural heritage, traditional patterns are at a critical point of inheritance and development in today's information society where artificial intelligence, Internet, VR and other technologies are developing rapidly. This study takes Chinese traditional textile patterns as the research object, and uses conditional generative adversarial network model and computer-aided technology to carry out research on the classification and generation of Chinese traditional textile patterns to solve the problems of old styles, lack of innovation and high cost of manual design of Chinese traditional textile patterns. We introduced deep learning conditional generative adversarial network model for pattern classification of Chinese traditional patterns; at the same time, we conducted theoretical research and innovative practice of conditional generative adversarial network model in the design of Chinese traditional patterns for cultural and creative purposes. After several training iterations, the conditional generative adversarial network model was able to follow the instructions to generate brand new patterns with traditional pattern features, generating brand new patterns with clear contours but pattern quality to be improved, with traditional textile pattern style features and clear pattern types, while generating brand new visual features that the patterns in the existing dataset do not have. The results of the design application of some patterns show that the designer can better realize the artificial intelligence collaborative design development of traditional textile patterns through the model, and the final design application works also present the artistic characteristics and cultural connotation of Chinese traditional textile patterns to a certain extent.