Spatial and Frequency Domain Adaptive Fusion Network for Image Deblurring

TL;DR

This paper introduces SFAFNet, a novel image deblurring network that adaptively fuses spatial and frequency domain features using a gated fusion block, leading to improved performance over existing methods.

Contribution

The paper proposes a spatial-frequency domain adaptive fusion network with a learnable low-pass filter and gated fusion mechanism, enabling effective integration of local and global features for deblurring.

Findings

Outperforms state-of-the-art methods on benchmark datasets.

Effectively captures both local details and global context.

Demonstrates robustness across various types of blur.

Abstract

Image deblurring aims to reconstruct a latent sharp image from its corresponding blurred one. Although existing methods have achieved good performance, most of them operate exclusively in either the spatial domain or the frequency domain, rarely exploring solutions that fuse both domains. In this paper, we propose a spatial-frequency domain adaptive fusion network (SFAFNet) to address this limitation. Specifically, we design a gated spatial-frequency domain feature fusion block (GSFFBlock), which consists of three key components: a spatial domain information module, a frequency domain information dynamic generation module (FDGM), and a gated fusion module (GFM). The spatial domain information module employs the NAFBlock to integrate local information. Meanwhile, in the FDGM, we design a learnable low-pass filter that dynamically decomposes features into separate frequency subbands, capturing the image-wide receptive field and enabling the adaptive exploration of global contextual information. Additionally, to facilitate information flow and the learning of complementary representations. In the GFM, we present a gating mechanism (GATE) to re-weight spatial and frequency domain features, which are then fused through the cross-attention mechanism (CAM). Experimental results demonstrate that our SFAFNet performs favorably compared to state-of-the-art approaches on commonly used benchmarks.