Application of AIGC Technology in Vernacular Architecture Design

In recent years, with the continuous development of large language models, artificial intelligence content automatic generation technology has made rapid progress and brought new challenges to the existing design paradigm. The constantly updated AIGC technology will change the future workflow. However, there are few researches on architectural image generation from the existing AIGC technology of architectural design. This study is based on the AIGC artificial intelligence tools such as midjoury and Stable Diffusion and the practice of vernacular architecture. How to use the instructions and methods of style customization and model deep learning to modify and optimize the original concept, so as to explore the potential of AIGC in improving the creativity and quality of architectural design. Through literature research and empirical research methods, this paper summarizes the influence of AIGC deep learning parameters on model training, and then proposes an optimization method for the generation of architectural scheme design process.