Canonical Makeup Transfer


Xinyu Lin (CUHK-Shenzhen), Kun Zhou (Shenzhen University), Xiaoguang Han (CUHK-Shenzhen), Jiangbo Lu (SmartMore Corporation)
The 35th British Machine Vision Conference
PDFPosterVideo (Right click to download)Supplementary

Abstract

Makeup transfer aims at adapting makeup styles from reference images to source non-makeup images, typically portraits. Despite significant progress in this area, challenges persist, particularly in effectively transferring makeup across faces with notable pose or expression variations. In this study, we introduce a canonical makeup transfer (CMT) approach to reduce the disparity between the source and reference style faces, while ensuring semantic consistency in the transformation process. By minimizing this transfer gap, our method facilitates robust and efficient interaction between source and style facial images. Additionally, we establish a comprehensive benchmark for evaluating the performance of state-of-the-art methods across diverse scenarios. Extensive experiments demonstrate that our approach achieves superior results in terms of transfer quality and artifact reduction.

Citation

@inproceedings{Lin_2025_BMVC,
author    = {Xinyu Lin and Kun Zhou and Xiaoguang Han and Jiangbo Lu},
title     = {Canonical Makeup Transfer},
booktitle = {36th British Machine Vision Conference 2025, {BMVC} 2025, Sheffield, UK, November 24-27, 2025},
publisher = {BMVA},
year      = {2025},
url       = {https://bmva-archive.org.uk/bmvc/2025/assets/papers/Paper_664/paper.pdf}
}


Copyright © 2025 The British Machine Vision Association and Society for Pattern Recognition
The British Machine Vision Conference is organised by The British Machine Vision Association and Society for Pattern Recognition. The Association is a Company limited by guarantee, No.2543446, and a non-profit-making body, registered in England and Wales as Charity No.1002307 (Registered Office: Dept. of Computer Science, Durham University, South Road, Durham, DH1 3LE, UK).

Imprint | Data Protection