바이오모픽 특성을 반영한 이미지 생성형 AI의 아트메이크업 실물 구현 연구

Abstract

Our study aims to propose and verify the industrial and educational application of a process that materializes images generated by image-generative AI. The process involves embodying biomorphic expression characteristics as art makeup on the three-dimensional human face. We categorized, on the basis of prior research, the concepts and artistic values of generative AI along with biomorphic design features—①dynamic morphology, ②color symbolism, ③ continuous connectivity, ④biological abstraction, and ⑤surrealism—to establish a theoretical foundation. In our study research, ChatGPT was used to derive biomorphic-specific keywords and prompts that were then applied to the image-generation platform Midjourney, which produced and selected one image for each characteristic (a total of five images). Based on the selected images, five works of art makeup were realized on actual human subjects along with addressing recurring challenges inherent in the digital-to-material translation process. Our findings demonstrate that generative AI effectively accelerates ideation and expands formal diversity. The proposed translation protocol ensures formal consistency and reproducibility among works. Ultimately, our study presents a systematic digital-to- material transformation framework that links generative AI, biomorphic aesthetics as well as art makeup to provide a foundation for the creative, educational, and practical advancement of beauty art practices.