Upstyle Design Inspired by Alphonse Mucha’s Works and Visualization Using Generative AI: Application of the RTF Framework

This study aims to reinterpret the formative characteristics of natural motifs in Alphonse Mucha’s Art Nouveau paintings into three-dimensional hair up-style designs, and to visualize the completed works through a generative AI system based on the RTF (Role–Task–Format) prompt framework, thereby proposing a practical convergence model that integrates analog hair artistry with digital image generation in beauty design. Five original works of Alphonse Mucha — Flowers: Iris (1898), Zodiac (1896), Laurel (1901), The Four Seasons: Autumn (1896), and Monaco Monte Carlo Poster (1897) — were selected based on Art Nouveau representativeness and convertibility into up-style structures. The form and digital color (RGB values extracted via Adobe Photoshop) of each work were analyzed and applied to five physical up-style works using roll, loop, and braid techniques. The completed works were then visualized through GPT-4o using prompts structured by the RTF framework, in which hair accessories, costumes, and backgrounds were generated while preserving the original hair form and color. As a result, first, the organic curves of natural motifs were successfully reinterpreted as linear S-curves and C-curves. Second, Art Nouveau’s voluminous elegance was reconstructed by integrating roll and loop techniques. Third, the AI visualization reduced approximately 80% of the direct cost otherwise required for physical costume and studio production. In conclusion, this study extends art-historical aesthetics into applied beauty design and demonstrates that generative AI functions as a cost-efficient and artistically expansive medium for beauty creative research.