GAN-based Face Mask Removal using Facial Landmarks and Pixel Errors in Masked Region

Hitoshi Yoshihashi, Naoto Ienaga, Maki Sugimoto

2022

Abstract

In 2020 and beyond, the opportunities to communicate with others while wearing a face mask have increased. A mask hides the mouth and facial muscles, making it difficult to convey facial expressions to others. In this study, we propose to use generative adversarial networks (GAN) to complete the facial region hidden by the mask. We defined custom loss functions that focus on the errors of the feature point coordinates of the face and the pixels in the masked region. As a result, we were able to generate images with higher quality than existing methods.

Paper Citation

in Harvard Style

Yoshihashi H., Ienaga N. and Sugimoto M. (2022). GAN-based Face Mask Removal using Facial Landmarks and Pixel Errors in Masked Region. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP; ISBN 978-989-758-555-5, SciTePress, pages 125-133. DOI: 10.5220/0010827500003124

in Bibtex Style

@conference{visapp22,
author={Hitoshi Yoshihashi and Naoto Ienaga and Maki Sugimoto},
title={GAN-based Face Mask Removal using Facial Landmarks and Pixel Errors in Masked Region},
booktitle={Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP},
year={2022},
pages={125-133},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010827500003124},
isbn={978-989-758-555-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP
TI - GAN-based Face Mask Removal using Facial Landmarks and Pixel Errors in Masked Region
SN - 978-989-758-555-5
AU - Yoshihashi H.
AU - Ienaga N.
AU - Sugimoto M.
PY - 2022
SP - 125
EP - 133
DO - 10.5220/0010827500003124
PB - SciTePress