Face Super Resolution Based on Identity Preserving V-Network

Numerous super-resolution methods have been developed to restore and upsample low-resolution and low-detail images to higher resolutions. Specifically, face super-resolution studies aim to restore various degradations in facial images while enhancing their resolution and preserving details. This stu...

Full description

Saved in:
Bibliographic Details
Main Authors: Hüseyin Eski, Ali Hüsameddin Ateş
Format: Article
Language:English
Published: Sakarya University 2025-03-01
Series:Sakarya University Journal of Computer and Information Sciences
Subjects:
Online Access:https://dergipark.org.tr/en/download/article-file/4110751
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849764510401822720
author Hüseyin Eski
Ali Hüsameddin Ateş
author_facet Hüseyin Eski
Ali Hüsameddin Ateş
author_sort Hüseyin Eski
collection DOAJ
description Numerous super-resolution methods have been developed to restore and upsample low-resolution and low-detail images to higher resolutions. Specifically, face super-resolution studies aim to restore various degradations in facial images while enhancing their resolution and preserving details. This study proposes the VNet architecture, which consists of a deep learning-based convolutional network for converting low-resolution and degraded facial images into high-quality and detailed images, and a pre-trained FaceNet model to preserve identity information. The architecture leverages the advantages of the Encoder-Decoder structure bidirectionally to maintain details and recover lost information. In the initial stage, the Encoder module compresses the image representation, filtering out unnecessary information. The Decoder module then reconstructs the high-resolution and restored image from the compressed representation. The use of residual connections in this process helps minimize information loss while preserving details. The final stage utilizes the identity feedback from the FaceNet model to enhance the image without deviating from the original identity context. Tests conducted on various facial datasets demonstrate that VNet achieves high metric performance in both super-resolution and restoration tasks. The results indicate that the proposed architecture is effective in producing realistic and high-quality versions of low-resolution and degraded facial images.
format Article
id doaj-art-76a09b04756e42e1bb776c05b252d1a2
institution DOAJ
issn 2636-8129
language English
publishDate 2025-03-01
publisher Sakarya University
record_format Article
series Sakarya University Journal of Computer and Information Sciences
spelling doaj-art-76a09b04756e42e1bb776c05b252d1a22025-08-20T03:05:07ZengSakarya UniversitySakarya University Journal of Computer and Information Sciences2636-81292025-03-0181273710.35377/saucis.8.91064.152535028Face Super Resolution Based on Identity Preserving V-NetworkHüseyin Eski0https://orcid.org/0000-0002-6006-3228Ali Hüsameddin Ateş1https://orcid.org/0000-0001-7690-7301SAKARYA UNIVERSITY, FACULTY OF COMPUTER AND INFORMATION SCIENCES, DEPARTMENT OF COMPUTER ENGINEERINGSAKARYA UNIVERSITY, FACULTY OF COMPUTER AND INFORMATION SCIENCES, DEPARTMENT OF COMPUTER ENGINEERINGNumerous super-resolution methods have been developed to restore and upsample low-resolution and low-detail images to higher resolutions. Specifically, face super-resolution studies aim to restore various degradations in facial images while enhancing their resolution and preserving details. This study proposes the VNet architecture, which consists of a deep learning-based convolutional network for converting low-resolution and degraded facial images into high-quality and detailed images, and a pre-trained FaceNet model to preserve identity information. The architecture leverages the advantages of the Encoder-Decoder structure bidirectionally to maintain details and recover lost information. In the initial stage, the Encoder module compresses the image representation, filtering out unnecessary information. The Decoder module then reconstructs the high-resolution and restored image from the compressed representation. The use of residual connections in this process helps minimize information loss while preserving details. The final stage utilizes the identity feedback from the FaceNet model to enhance the image without deviating from the original identity context. Tests conducted on various facial datasets demonstrate that VNet achieves high metric performance in both super-resolution and restoration tasks. The results indicate that the proposed architecture is effective in producing realistic and high-quality versions of low-resolution and degraded facial images.https://dergipark.org.tr/en/download/article-file/4110751face super resolutionface restorationsuper resolutiondeep learning
spellingShingle Hüseyin Eski
Ali Hüsameddin Ateş
Face Super Resolution Based on Identity Preserving V-Network
Sakarya University Journal of Computer and Information Sciences
face super resolution
face restoration
super resolution
deep learning
title Face Super Resolution Based on Identity Preserving V-Network
title_full Face Super Resolution Based on Identity Preserving V-Network
title_fullStr Face Super Resolution Based on Identity Preserving V-Network
title_full_unstemmed Face Super Resolution Based on Identity Preserving V-Network
title_short Face Super Resolution Based on Identity Preserving V-Network
title_sort face super resolution based on identity preserving v network
topic face super resolution
face restoration
super resolution
deep learning
url https://dergipark.org.tr/en/download/article-file/4110751
work_keys_str_mv AT huseyineski facesuperresolutionbasedonidentitypreservingvnetwork
AT alihusameddinates facesuperresolutionbasedonidentitypreservingvnetwork