Spatial-based generative adversarial network for makeup transfer
L. -Y. Ke, Weiyu Lan, H. -Y. Lin, Chih‐Hsien Hsia · IET conference proceedings. · 2024
In recent years, with the rise of social media platforms, there has been a growing emphasis on personal appearance, leading to an increased demand for makeup products. However, finding the suitable makeup can be time-consuming, constant experimentation with different products can potentially have long-term effects on the skin. Therefore, this study proposes a model architecture for makeup transfer, aiming to accurately transform partial makeup using the exact feature distribution matching (EFDM) method. As results, demonstrate that the proposed model outperforms existing research in terms of both quality and quantitative comparison. Additionally, this model's frechet inception distance (FID) achieves 38.52.