An improved method for partial occlusion face inpainting

Qingyu Liu; Roben A. Juanatas

doi:10.1117/12.2684614

19 October 2023 An improved method for partial occlusion face inpainting

Qingyu Liu, Roben A. Juanatas

Proceedings Volume 12709, Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023); 1270948 (2023) https://doi.org/10.1117/12.2684614
Event: Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023), 2023, Nanjing, China

Abstract

Face with mask would lead to a reduction in the number of features that can be extracted from the face, thus severely reducing the accuracy of face recognition. Image inpainting methods can be used to recover the obscured regions. A face image inpainting model based on pix2pix generative adversarial network is proposed for the current blurring problem after face image inpainting. The generator network and discriminator network in the model are trained adversarially to achieve the restoration of obscured faces. Multi-branch residual modules and dilated convolutional layers are added to the generator network as a way to guarantee the boundary consistency of the restored images. In addition, a face dataset wearing masks is constructed on the CelebA dataset. According to the experimental findings, the inpainting result is effective, and both PSNR and SSIM are much better than those of the original pix2pix and other methods.

Citation Download Citation

Qingyu Liu and Roben A. Juanatas "An improved method for partial occlusion face inpainting", Proc. SPIE 12709, Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023), 1270948 (19 October 2023); https://doi.org/10.1117/12.2684614

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
9 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Feature extraction

Education and training

Image restoration

Convolution

Facial recognition systems

Data modeling

Image quality

Show All Keywords

Keywords/Phrases

Search In:

Publication Years