Papers › LoFormer: Local Frequency Transformer for Image Deblurring

LoFormer: Local Frequency Transformer for Image Deblurring

24 Jul 2024arXiv:2407.16993archive 2025-07-28

Xintian Mao, Jiansheng Wang, Xingran Xie, Qingli Li, Yan Wang

Due to the computational complexity of self-attention (SA), prevalent techniques for image deblurring often resort to either adopting localized SA or employing coarse-grained global SA methods, both of which exhibit drawbacks such as compromising global modeling or lacking fine-grained correlation. In order to address this issue by effectively modeling long-range dependencies without sacrificing fine-grained details, we introduce a novel approach termed Local Frequency Transformer (LoFormer). Within each unit of LoFormer, we incorporate a Local Channel-wise SA in the frequency domain (Freq-LC) to simultaneously capture cross-covariance within low- and high-frequency local windows. These operations offer the advantage of (1) ensuring equitable learning opportunities for both coarse-grained structures and fine-grained details, and (2) exploring a broader range of representational properties compared to coarse-grained global SA methods. Additionally, we introduce an MLP Gating mechanism complementary to Freq-LC, which serves to filter out irrelevant features while enhancing global learning capabilities. Our experiments demonstrate that LoFormer significantly improves performance in the image deblurring task, achieving a PSNR of 34.09 dB on the GoPro dataset with 126G FLOPs. https://github.com/DeepMed-Lab-ECNU/Single-Image-Deblur

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

deepmed-lab-ecnu/single-image-deblur officialmentioned in papermentioned on GitHubpytorch report
INVOKERer/LoFormer mentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DeblurringImage Deblurring

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Deblurring HIDE (trained on GOPRO) LoFormer PSNR (sRGB) 31.86 #5 of 26 Archive leaderboard report
Deblurring HIDE (trained on GOPRO) LoFormer SSIM (sRGB) 0.949 #5 of 26 Archive leaderboard report
Deblurring RealBlur-J LoFormer PSNR (sRGB) 32.90 #6 of 17 Archive leaderboard report
Deblurring RealBlur-J LoFormer SSIM (sRGB) 0.933 #6 of 17 Archive leaderboard report
Deblurring RealBlur-R LoFormer PSNR (sRGB) 40.23 #7 of 17 Archive leaderboard report
Deblurring RealBlur-R LoFormer SSIM (sRGB) 0.974 #7 of 17 Archive leaderboard report
Image Deblurring GoPro LoFormer PSNR 34.09 #5 of 55 Archive leaderboard report
Image Deblurring GoPro LoFormer SSIM 0.969 #5 of 55 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections