Face attribute editing based on self-attention mechanism and generative adversarial network

Naen Xu; Wenxin Yao

doi:10.1117/12.3010222

20 October 2023 Face attribute editing based on self-attention mechanism and generative adversarial network

Naen Xu, Wenxin Yao

Proceedings Volume 12814, Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023); 1281411 (2023) https://doi.org/10.1117/12.3010222
Event: Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023), 2023, Chongqing, China

Abstract

The purpose of face attribute editing is to control the specific attributes in the face image according to the specified face attribute tags, and obtain a synthetic face image that meets the specific attributes. In recent years, face attribute editing has been widely used in the fields of beauty, entertainment and face recognition. Based on the problem of inaccurate attribute editing in the current model, this paper proposes a face attribute editing method based on self-attention mechanism and generative adversarial network. Taking selective transfer generative adversarial network (STGAN) as the basic model, a self-attention mechanism is added after the convolutional layer of the decoder. At the same time, the loss function is improved, and the latent space encoding consistency loss is added to the original objective function. Experiments were carried out on the CelebA face dataset, and it was found that the model has improved compared with StarGAN, AttGAN and STGAN in terms of image generation quality and face attribute editing accuracy.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Naen Xu and Wenxin Yao "Face attribute editing based on self-attention mechanism and generative adversarial network", Proc. SPIE 12814, Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023), 1281411 (20 October 2023); https://doi.org/10.1117/12.3010222

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
7 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Image quality

Image restoration

Convolution

Education and training

Eye models

Gallium nitride

Image compression

Show All Keywords

Keywords/Phrases

Search In:

Publication Years