DynArt ChatGPT: a Platform for the Generation of Dynamic Non-Traditional New Year Paintings

Wang Hongliang,Zhou Manling,Zhang Jiayi,Jin Cong,Wang Jing
DOI: https://doi.org/10.1109/iccsi58851.2023.10304019
2023-01-01
Abstract:In a modern, technology-driven society, the preservation and development of China’s Intangible Cultural Heritage (ICH) has encountered three major contradictions: economic development, technological renewal and cultural change, resulting in a lack of people to pass on the heritage, difficulties in innovation and inadequate protection. In order to promote and develop Chinese traditional art, we had a vision: can we combine the conversation function of ChatGPT with the traditional art of ICH? Based on this idea, we have built a system - DynaArt ChatGPT. DynaArt ChatGPT can extract the keywords from the sample given by ChatGPT and then generate a description related to the sample based on the keywords, we input the description into Lumen5 model, Lumen5 will generate the corresponding dynamic video according to the description we input, after the experiment we think the dynamic painting generation After experiments we believe that the system can provide a new interpretation of some familiar Chinese folk tales.
What problem does this paper attempt to address?