Blind Image Deconvolution by Generative-based Kernel Prior and Initializer via Latent Encoding

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhang, Jiangtao, Yue, Zongsheng, Wang, Hui, Zhao, Qian, Meng, Deyu
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916331361665024
author Zhang, Jiangtao
Yue, Zongsheng
Wang, Hui
Zhao, Qian
Meng, Deyu
author_facet Zhang, Jiangtao
Yue, Zongsheng
Wang, Hui
Zhao, Qian
Meng, Deyu
contents Blind image deconvolution (BID) is a classic yet challenging problem in the field of image processing. Recent advances in deep image prior (DIP) have motivated a series of DIP-based approaches, demonstrating remarkable success in BID. However, due to the high non-convexity of the inherent optimization process, these methods are notorious for their sensitivity to the initialized kernel. To alleviate this issue and further improve their performance, we propose a new framework for BID that better considers the prior modeling and the initialization for blur kernels, leveraging a deep generative model. The proposed approach pre-trains a generative adversarial network-based kernel generator that aptly characterizes the kernel priors and a kernel initializer that facilitates a well-informed initialization for the blur kernel through latent space encoding. With the pre-trained kernel generator and initializer, one can obtain a high-quality initialization of the blur kernel, and enable optimization within a compact latent kernel manifold. Such a framework results in an evident performance improvement over existing DIP-based BID methods. Extensive experiments on different datasets demonstrate the effectiveness of the proposed method.
format Preprint
id arxiv_https___arxiv_org_abs_2407_14816
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Blind Image Deconvolution by Generative-based Kernel Prior and Initializer via Latent Encoding
Zhang, Jiangtao
Yue, Zongsheng
Wang, Hui
Zhao, Qian
Meng, Deyu
Computer Vision and Pattern Recognition
I.4.4
Blind image deconvolution (BID) is a classic yet challenging problem in the field of image processing. Recent advances in deep image prior (DIP) have motivated a series of DIP-based approaches, demonstrating remarkable success in BID. However, due to the high non-convexity of the inherent optimization process, these methods are notorious for their sensitivity to the initialized kernel. To alleviate this issue and further improve their performance, we propose a new framework for BID that better considers the prior modeling and the initialization for blur kernels, leveraging a deep generative model. The proposed approach pre-trains a generative adversarial network-based kernel generator that aptly characterizes the kernel priors and a kernel initializer that facilitates a well-informed initialization for the blur kernel through latent space encoding. With the pre-trained kernel generator and initializer, one can obtain a high-quality initialization of the blur kernel, and enable optimization within a compact latent kernel manifold. Such a framework results in an evident performance improvement over existing DIP-based BID methods. Extensive experiments on different datasets demonstrate the effectiveness of the proposed method.
title Blind Image Deconvolution by Generative-based Kernel Prior and Initializer via Latent Encoding
topic Computer Vision and Pattern Recognition
I.4.4
url https://arxiv.org/abs/2407.14816