| Title |
An Approach to Utilizing Generative AI for Spatial Design Visualization based on Space Identity |
| Authors |
신재영(Shin, Jaeyoung) ; 이진국(Lee, Jin-Kook) |
| DOI |
http://doi.org/10.14774/JKIID.2023.32.6.026 |
| Keywords |
Generative Artificial Intelligence; Space Identity; Spatial Design Visualization; Image Generation |
| Abstract |
Numerous companies have embraced space marketing strategies as a means to cultivate a unique brand identity and image. The design of interior spaces has emerged as a pivotal role in conveying the brand image by crafting a distinct spatial identity. Traditionally, designers conceptualize spatial identity and subsequently iterate through the process of generating and visualizing a multitude of design alternatives to ultimately arrive at the optimal spatial design. The advent of Generative AI has exhibited exceptional capabilities in the realm of image creation and is poised to significantly enhance the efficiency of creative design work. This paper introduces an innovative approach to visualizing interior space design that captures the elements of spatial identity through the utilization of image generation AI technology. In pursuit of this objective, we have developed a spatial identity visualization model with a range of materials, shapes/patterns, and colors, utilizing real-world interior design cases as training data source.
Furthermore, we present diverse application methodologies for leveraging this trained model, offering practical insights through empirical cases across three distinct scenarios. The proposed approach is expected to contribute to streamlining the early phase of design by facilitating the generation of reference images based on the designated spatial identity. In addition, this enables designers to swiftly explore customized space design alternatives and develop imaginative space designs, thereby reinforcing the spatial identity. |