PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance

Reference-based blind face restoration (RefBFR) has gained considerable attention because it utilizes additional reference images to restore facial images in situations where the degradation is caused by unknown factors, making it particularly useful in real-world applications. Recently, guided diff...

Full description

Saved in:
Bibliographic Details
Main Authors: Geon Min, Tae Bok Lee, Yong Seok Heo
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/24/22/7112
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846152470620798976
author Geon Min
Tae Bok Lee
Yong Seok Heo
author_facet Geon Min
Tae Bok Lee
Yong Seok Heo
author_sort Geon Min
collection DOAJ
description Reference-based blind face restoration (RefBFR) has gained considerable attention because it utilizes additional reference images to restore facial images in situations where the degradation is caused by unknown factors, making it particularly useful in real-world applications. Recently, guided diffusion models have demonstrated exceptional performance in this task without requiring training. They achieve this by integrating gradients of the losses where each loss reflects the different desired properties of the additional external images. However, these approaches fail to consider potential conflicts between gradients of multiple losses, which can lead to sub-optimal results. To address this issue, we introduce Pivot Direction Gradient guidance (PDGrad), a novel gradient adjustment method for RefBFR within a guided diffusion framework. To this end, we first define the loss function based on both low-level and high-level features. For loss at each feature level, both the coarsely restored image and the reference image are fully integrated. In cases of conflicting gradients, a pivot gradient is established for each level and other gradients are aligned to it, ensuring that the strengths of both images are maximized. Additionally, if the magnitude of the adjusted gradient exceeds that of the pivot gradient, it is adaptively scaled according to the ratio between the two, placing greater emphasis on the pivot. Extensive experimental results on the CelebRef-HQ dataset show that the proposed PDGrad significantly outperforms competitive approaches both quantitatively and qualitatively.
format Article
id doaj-art-10af5f93f38a4fabb26d1bd1a3c7fc2c
institution Kabale University
issn 1424-8220
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-10af5f93f38a4fabb26d1bd1a3c7fc2c2024-11-26T18:20:45ZengMDPI AGSensors1424-82202024-11-012422711210.3390/s24227112PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient GuidanceGeon Min0Tae Bok Lee1Yong Seok Heo2Department of Artificial Intelligence, Ajou University, Suwon 16499, Republic of KoreaDepartment of Artificial Intelligence, Ajou University, Suwon 16499, Republic of KoreaDepartment of Artificial Intelligence, Ajou University, Suwon 16499, Republic of KoreaReference-based blind face restoration (RefBFR) has gained considerable attention because it utilizes additional reference images to restore facial images in situations where the degradation is caused by unknown factors, making it particularly useful in real-world applications. Recently, guided diffusion models have demonstrated exceptional performance in this task without requiring training. They achieve this by integrating gradients of the losses where each loss reflects the different desired properties of the additional external images. However, these approaches fail to consider potential conflicts between gradients of multiple losses, which can lead to sub-optimal results. To address this issue, we introduce Pivot Direction Gradient guidance (PDGrad), a novel gradient adjustment method for RefBFR within a guided diffusion framework. To this end, we first define the loss function based on both low-level and high-level features. For loss at each feature level, both the coarsely restored image and the reference image are fully integrated. In cases of conflicting gradients, a pivot gradient is established for each level and other gradients are aligned to it, ensuring that the strengths of both images are maximized. Additionally, if the magnitude of the adjusted gradient exceeds that of the pivot gradient, it is adaptively scaled according to the ratio between the two, placing greater emphasis on the pivot. Extensive experimental results on the CelebRef-HQ dataset show that the proposed PDGrad significantly outperforms competitive approaches both quantitatively and qualitatively.https://www.mdpi.com/1424-8220/24/22/7112reference-based blind face restorationclassifier guidance diffusion modelconflicting gradients
spellingShingle Geon Min
Tae Bok Lee
Yong Seok Heo
PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance
Sensors
reference-based blind face restoration
classifier guidance diffusion model
conflicting gradients
title PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance
title_full PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance
title_fullStr PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance
title_full_unstemmed PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance
title_short PDGrad: Guiding Diffusion Model for Reference-Based Blind Face Restoration with Pivot Direction Gradient Guidance
title_sort pdgrad guiding diffusion model for reference based blind face restoration with pivot direction gradient guidance
topic reference-based blind face restoration
classifier guidance diffusion model
conflicting gradients
url https://www.mdpi.com/1424-8220/24/22/7112
work_keys_str_mv AT geonmin pdgradguidingdiffusionmodelforreferencebasedblindfacerestorationwithpivotdirectiongradientguidance
AT taeboklee pdgradguidingdiffusionmodelforreferencebasedblindfacerestorationwithpivotdirectiongradientguidance
AT yongseokheo pdgradguidingdiffusionmodelforreferencebasedblindfacerestorationwithpivotdirectiongradientguidance