Frequency-domain Learning with Kernel Prior for Blind Image Deblurring
Jixiang Sun, Fei Lei, Jiawei Zhang, Wenxiu Sun, Yujiu Yang
TL;DR
This work tackles blind image deblurring by incorporating explicit blur kernel priors into a frequency-domain Transformer. It introduces a Kernel Estimation Module to predict per-pixel kernels and a Frequency Integration Module that fuses kernel information in the frequency domain through a novel Frequency Attention mechanism, augmented by a multi-scale encoder–decoder integration. The approach yields state-of-the-art results on GoPro, HIDE, and RealBlur benchmarks and demonstrates robust generalization to out-of-domain data, albeit with higher computational cost. Overall, the combination of kernel priors and frequency-domain fusion offers a practical path toward more generalizable deblurring models with explicit degradation information.
Abstract
While achieving excellent results on various datasets, many deep learning methods for image deblurring suffer from limited generalization capabilities with out-of-domain data. This limitation is likely caused by their dependence on certain domain-specific datasets. To address this challenge, we argue that it is necessary to introduce the kernel prior into deep learning methods, as the kernel prior remains independent of the image context. For effective fusion of kernel prior information, we adopt a rational implementation method inspired by traditional deblurring algorithms that perform deconvolution in the frequency domain. We propose a module called Frequency Integration Module (FIM) for fusing the kernel prior and combine it with a frequency-based deblurring Transfomer network. Experimental results demonstrate that our method outperforms state-of-the-art methods on multiple blind image deblurring tasks, showcasing robust generalization abilities. Source code will be available soon.