SR-R$^2$KAC: Improving Single Image Defocus Deblurring

TL;DR

This paper introduces SR-R$^2$KAC, a novel deep learning approach that effectively handles large spatially varying defocus blurs in images by recursively simulating large inverse kernels and leveraging multi-scale information.

Contribution

The paper proposes R$^2$KAC with recursive atrous convolutions and residual shortcuts, significantly improving defocus deblurring performance over existing inverse kernel methods.

Findings

Achieves state-of-the-art results on defocus deblurring benchmarks.

Effectively handles large, spatially varying defocus blurs.

Reduces ringing artifacts through residual connections.

Abstract

We propose an efficient deep learning method for single image defocus deblurring (SIDD) by further exploring inverse kernel properties. Although the current inverse kernel method, i.e., kernel-sharing parallel atrous convolution (KPAC), can address spatially varying defocus blurs, it has difficulty in handling large blurs of this kind. To tackle this issue, we propose a Residual and Recursive Kernel-sharing Atrous Convolution (R$^2$KAC). R$^2$KAC builds on a significant observation of inverse kernels, that is, successive use of inverse-kernel-based deconvolutions with fixed size helps remove unexpected large blurs but produces ringing artifacts. Specifically, on top of kernel-sharing atrous convolutions used to simulate multi-scale inverse kernels, R$^2$KAC applies atrous convolutions recursively to simulate a large inverse kernel. Specifically, on top of kernel-sharing atrous convolutions, R$^2$KAC stacks atrous convolutions recursively to simulate a large inverse kernel. To further alleviate the contingent effect of recursive stacking, i.e., ringing artifacts, we add identity shortcuts between atrous convolutions to simulate residual deconvolutions. Lastly, a scale recurrent module is embedded in the R$^2$KAC network, leading to SR-R$^2$KAC, so that multi-scale information from coarse to fine is exploited to progressively remove the spatially varying defocus blurs. Extensive experimental results show that our method achieves the state-of-the-art performance.