A Study on the Creation of Lip Makeup Images Based on Generative AI: Focusing on Techniques, Textures, and Gradient Elements

Recently, with the growing presence of generative AI across industries, its potential in the beauty sector is also gaining attention. However, the capability of generative AI to accurately reproduce detailed aspects of actual makeup expression has yet to be fully examined. To address this, the present study focused on lip makeup as a representative element to evaluate the precision of generative AI. Lip samples were generated, and both independent and combined experiments were conducted based on three elements: technique, texture, and gradient. The generated images were evaluated using predefined criteria. In the independent experiments, texture showed high reproducibility, whereas implementing certain technique features such as incurve revealed limitations. In combined experiments, conflicts among elements further reduced the representation of technique and gradient, and some prompts were found to be rejected due to generation policies. These findings suggest that generative AI still faces technical challenges in achieving detailed makeup expression and integrating complex elements. To expand its applicability in the beauty industry, future efforts should focus on refining prompt design, developing technical post-processing methods, and improving platform policies in parallel.