Application Design and Analysis Method of Dialect Cultural Symbols for Multimodal Large Model

Abstract

At present, multimodal large model with the ability to quickly integrate text, image, audio and other multi-dimensional data has become the forefront of the research field. Benefiting from multi-modal technology and AI large model, the design industry has also achieved excellent results. However, the complexity and diversity of cultural symbols in dialects challenge traditional design methods. The integration of multi-modal information and the optimization of creative design with new quality productivity will effectively drive the modern dissemination of dialect cultural symbols. This paper constructs a theoretical framework of multimodal model and dialect culture, covering three levels: cross-modal semantic understanding, generative creation and personalized adaptation. For the first time, this paper applies the multi-modal large model to the design of dialect cultural symbols, analyzes the resources of Nanning dialect, and puts forward the idea of cross-modal alignment and dynamic adaptation mechanism to realize the accurate extraction and creative design of dialect cultural symbols. With the help of the multi-modal large model, the recognition accuracy and emotional expression of dialect cultural symbols will be significantly improved, and the design scheme will be generated by the use of digital and intelligent technology, which can not only enrich the theoretical system of multi-modal intelligent design, but also provide a new path for regional cultural inheritance and design innovation, and effectively solve the industry pain point of the separation of cultural connotation and functional form in traditional cultural and creative design.