BFRFormer: Transformer-based generator for Real-World Blind Face Restoration

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ge, Guojing, Song, Qi, Zhu, Guibo, Zhang, Yuting, Chen, Jinglu, Xin, Miao, Tang, Ming, Wang, Jinqiao
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914696351711232
author Ge, Guojing
Song, Qi
Zhu, Guibo
Zhang, Yuting
Chen, Jinglu
Xin, Miao
Tang, Ming
Wang, Jinqiao
author_facet Ge, Guojing
Song, Qi
Zhu, Guibo
Zhang, Yuting
Chen, Jinglu
Xin, Miao
Tang, Ming
Wang, Jinqiao
contents Blind face restoration is a challenging task due to the unknown and complex degradation. Although face prior-based methods and reference-based methods have recently demonstrated high-quality results, the restored images tend to contain over-smoothed results and lose identity-preserved details when the degradation is severe. It is observed that this is attributed to short-range dependencies, the intrinsic limitation of convolutional neural networks. To model long-range dependencies, we propose a Transformer-based blind face restoration method, named BFRFormer, to reconstruct images with more identity-preserved details in an end-to-end manner. In BFRFormer, to remove blocking artifacts, the wavelet discriminator and aggregated attention module are developed, and spectral normalization and balanced consistency regulation are adaptively applied to address the training instability and over-fitting problem, respectively. Extensive experiments show that our method outperforms state-of-the-art methods on a synthetic dataset and four real-world datasets. The source code, Casia-Test dataset, and pre-trained models are released at https://github.com/s8Znk/BFRFormer.
format Preprint
id arxiv_https___arxiv_org_abs_2402_18811
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle BFRFormer: Transformer-based generator for Real-World Blind Face Restoration
Ge, Guojing
Song, Qi
Zhu, Guibo
Zhang, Yuting
Chen, Jinglu
Xin, Miao
Tang, Ming
Wang, Jinqiao
Computer Vision and Pattern Recognition
Blind face restoration is a challenging task due to the unknown and complex degradation. Although face prior-based methods and reference-based methods have recently demonstrated high-quality results, the restored images tend to contain over-smoothed results and lose identity-preserved details when the degradation is severe. It is observed that this is attributed to short-range dependencies, the intrinsic limitation of convolutional neural networks. To model long-range dependencies, we propose a Transformer-based blind face restoration method, named BFRFormer, to reconstruct images with more identity-preserved details in an end-to-end manner. In BFRFormer, to remove blocking artifacts, the wavelet discriminator and aggregated attention module are developed, and spectral normalization and balanced consistency regulation are adaptively applied to address the training instability and over-fitting problem, respectively. Extensive experiments show that our method outperforms state-of-the-art methods on a synthetic dataset and four real-world datasets. The source code, Casia-Test dataset, and pre-trained models are released at https://github.com/s8Znk/BFRFormer.
title BFRFormer: Transformer-based generator for Real-World Blind Face Restoration
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2402.18811