UniDeblur: Unified Motion Deblurring via Fourier Convolution Learning

Xintian Mao; Jiansheng Wang; Huaiyuan Wang; Qingli Li; Yan Wang

cs.CVOct 9, 2026

UniDeblur: Unified Motion Deblurring via Fourier Convolution Learning

Authors: Xintian Mao, Jiansheng Wang, Huaiyuan Wang, Qingli Li, Yan Wang

Organizations: Shanghai Key Laboratory of Multidimensional Information Processing, East China Normal University, China. · Shanghai Institute of Technical Physics, Chinese Academy of Sciences, China.

Abstract

Image deblurring is a fundamental task in computer vision, yet existing methods mainly specialize in isolated scenarios. Consequently, the deblurring performance of these models drops significantly across scenarios with different cameras and image contents. This motivates the development of a unified motion deblurring framework that can handle diverse blur patterns. To this end, we construct a multi-source dataset and develop a data pruning and quality control pipeline. Then, to restore blur images with various motion blur kernels across different scenarios, we propose a unified motion deblurring framework via Fourier Convolution Learning. Existing deblurring networks usually rely on pixel-level mappings and overlook the convolutional nature of blur. Thus, the network mostly fits the image content itself rather than learning the convolution process, making it difficult to handle multi-source blur images. Instead of relying on spatial mapping, we perform blur-adaptive global filter learning in Fourier space, which avoids the limited spatial support of kernel estimation. Extensive experiments demonstrate that our unified model effectively handles motion blurs across diverse sources and scenes, achieving competitive performance across both in-domain and OOD benchmarks. Code has been released at github.com/DeepMed-Lab-ECNU/Single-Image-Deblur.

Figures & tables

Figure 1: UniDeblur via Fourier Convolution Learning.

Figure 2: Visualization of blur images, deconvolution kernels and kernels by FCL on various datasets.

Dataset	Acquisition	Images	Videos	Camera	Resolution	Content
GoPro Nah et al. (2017)	Averaged	3.2K	33	GoPro HERO4	1280 $\times$ 720	Outdoor Scenes
HIDE Shen et al. (2019)	Averaged	8.4K	31	GoPro HERO	1280 $\times$ 720	Pedestrian in Outdoor
REDS Nah et al. (2021)	Averaged	30K	300	GoPro HERO6	1280 $\times$ 720	Outdoor Scenes
RSBlur Rim et al. (2022)	Beam Splitter	13.3K	697	-	$\approx$ 1915 $\times$ 1200	Outdoor Scenes
RealBlur Rim et al. (2020)	Beam Splitter	9.1K	232	Sony A7RM3	$\approx$ 660 $\times$ 750	Static Object at Night
LOLBlur Zhou et al. (2022)	Simulated	12K	200	Sony RX10 IV	1120 $\times$ 640	Low-light Scenes

Table 1: Summary of various motion deblurring datasets.

Figure 3: Architecture of UniDeblur, Fourier Convolution Learning and Basicblocks.

Figure 4: Various flaws in OOD datasets.

	Train			Test
Dataset	Images	Patches	Resolution	Images	Removed	Resolution	Step
GoPro Nah et al. (2017)	2,103	16,824	512 $\times$ 512	1,111	0	1280 $\times$ 720	1
HIDE Shen et al. (2019)	6,397	12,794	720 $\times$ 720	203	0	1280 $\times$ 720	10
REDS Nah et al. (2021)	24,000	48,000	720 $\times$ 720	299	1	1280 $\times$ 720	10
RSBlur Rim et al. (2022)	8,878	53,268	720 $\times$ 720	336	0	$\approx$ 1915 $\times$ 1200	10
RealBlur Rim et al. (2020)	7,516	15,032	$\approx$ 660 $\times$ 750	1,960	0	640 $\times$ 640	1

Table 2: Summary of training and testing datasets.

	GoPro			HIDE			REDS			RSBlur			RealBlur		LOLBlur
Method	PSNR	SSIM	MUSIQ	PSNR	SSIM	MUSIQ	PSNR	SSIM	MUSIQ	PSNR	SSIM	MUSIQ	PSNR	SSIM	PSNR	SSIM
GT	-	1.000	47.97	-	1.000	60.43	-	1.000	66.37	-	1.000	37.31	-	1.000	-	1.000
Blur	25.64	0.793	24.87	23.94	0.759	26.28	26.34	0.772	39.01	28.74	0.850	21.90	30.08	0.874	33.82	0.904
MIMOUNet	30.47	0.906	39.91	28.93	0.888	47.22	32.14	0.908	62.08	32.30	0.917	30.57	33.01	0.925	41.70	0.977
DeepRFT	31.92	0.932	41.11	30.84	0.921	51.70	32.36	0.912	61.21	33.45	0.937	32.40	33.79	0.938	43.38	0.982
NAFNet	32.81	0.940	43.61	31.11	0.924	54.99	33.66	0.931	64.50	33.89	0.941	34.62	33.93	0.939	44.22	0.985

Table 3: Comparisons on in-domain datasets.

	MIORe		VAR-MIORe		BSD1		BSD2		BSD3		FoundIR		FLOPs	Mem
Method	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	(G)	(GB)
GT	0.3076	55.70	0.3045	56.81	0.2695	55.05	0.2919	51.42	0.2658	49.70	0.3207	61.42	-	-
Blur	0.2120	36.75	0.1953	26.85	0.2097	40.35	0.1854	34.15	0.1740	30.89	0.1477	23.76	-	-
MIMOUNet	0.2422	44.41	0.2147	32.97	0.2415	49.08	0.2306	42.24	0.2111	41.53	0.2756	48.21	67	12
DeepRFT	0.2797	52.01	0.2299	39.91	0.2656	52.70	0.2787	48.23	0.2395	47.76	0.3146	53.90	80	24
NAFNet	0.2704	51.00	0.2224	37.88	0.2886	53.74	0.2923	49.28	0.2649	48.02	0.3102	53.82	63	35

Table 4: Comparisons on OOD datasets.

Figure 5: Visual comparison on GoPro, RSBlur, RealBlur, MIORe and VAR-MIORe.

	Backbone			Fourier Convolution Learning					GoPro	HIDE	RSBlur	LOLBlur	FLOPs	Mem
Exp	Basicblock	Level1 $\sim$ 3	Dim	Basicblock	mode	MS	Num	Dim	PSNR	PSNR	PSNR	PSNR	(G)	(GB)
1	NAFBlock	[1, 4, 8]	32	-	-	-	-	-	31.96	30.27	33.26	43.32	14.20	21
2	NAFBlock	[1, 4, 8]	32	ConvBlock	$\oplus$	$\times$	4	48	32.08	30.41	33.35	43.49	15.20	22
3	NAFBlock	[1, 4, 8]	32	ConvBlock	$\odot$	$\times$	4	48	32.23	30.56	33.51	43.62	15.20	22
4	NAFBlock	[1, 4, 8]	32	MLPBlock	$\odot$	$\times$	4	48	32.27	30.59	33.52	43.61	15.36	22
5	NAFBlock	[1, 4, 8]	32	GatedBlock	$\oplus$	$\times$	4	48	32.23	30.57	33.51	43.63	15.52	22

Table 5: Ablation for Fourier Convolution Learning. Exp is the index of each experiment. Dim is the channel number of Level 1. Num means the number of basicblock of FCL. Mode means get

\mathcal{F}(\mathrm{x}_{out})

via

\odot

(multiplication) or

\oplus

(addition). MS denotes applying FCL in multi-scale or not. Exp 6 and 13 are trained without auxiliary supervision.

Figure 6: Visualization of the frequency filters $\mathcal{F}(\mathrm{k})$ and corresponding spatial kernels $\mathrm{k}$ at Levels 1 $\sim$ 3, together with the averaged $\mathrm{x}_{in}~/~\mathrm{x}_{out}$ and the corresponding $\mathrm{B}~/~\mathrm{S}~/~\mathrm{k}_{blur}~/~\mathrm{k}_{deblur}$ on GoPro, RSBlur, and VAR-MIORe.

Appendix figures & tables18 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 7: Visualization of the frequency filters $\mathcal{F}(\mathrm{k})$ and corresponding spatial kernels $\mathrm{k}$ at Levels 1 $\sim$ 3, together with the averaged $\mathrm{x}_{in}~/~\mathrm{x}_{out}$ and the corresponding $\mathrm{B}~/~\mathrm{S}~/~\mathrm{k}_{blur}~/~\mathrm{k}_{deblur}$ on HIDE, REDS, LOLBlur, BSD and FoundIR.

Figure 8: Color shift and residual blur of ReLoBlur dataset.

	GT		Blur		PromptIR		MoCEIR		EVSSM		UniDeblur-L
Dataset	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ	CLIPIQA	MUSIQ
ReLoBlur	0.2103	40.14	0.1889	37.20	0.2546	41.30	0.2169	39.85	0.2841	43.61	0.2913	43.74
RealLOL	-	-	0.1370	26.12	0.1979	29.74	0.2028	29.69	0.1995	30.04	0.2088	30.28

Appendix

Table 6: Comparisons on ReLoBlur and RealLOL.

	GoPro	HIDE	RealBlur	LOLBlur	REDS	RSBlur
train	PSNR	PSNR	PSNR	PSNR	PSNR	PSNR
GoPro	32.60	30.44	30.04	35.24	27.08	26.22
GoPro+HIDE	32.88	31.11	29.99	34.94	27.15	28.62
GoPro+HIDE+RealBlur	32.53	30.80	33.67	37.60	27.81	30.10
GoPro+HIDE+LOLBlur	32.63	30.93	30.63	43.46	28.58	28.68
GoPro+HIDE+RealBlur+LOLBlur	32.45	30.75	33.72	43.28	29.62	30.31

Appendix

Table 7: Performance of NAFNet32 trained by different datasets .

Figure 9: Visualization comparisons on GoPro dataset.

Figure 10: Visualization comparisons on HIDE dataset.

Figure 11: Visualization comparisons on REDS dataset.

Figure 12: Visualization comparisons on RSBlur dataset.

Figure 13: Visualization comparisons on RealBlur and LOLBlur dataset.

Figure 14: Visualization comparisons on MIORe and VAR-MIORe dataset.

Figure 15: Visualization comparisons on BSD dataset.

Figure 16: Visualization comparisons on FoundIR dataset.

Figure 17: Visualization comparisons on ReLoBlur dataset.

Figure 18: Deleted image from REDS dataset.

Figure 19: Deleted image from BSD2 dataset.

Figure 20: Deleted image from BSD3 dataset.

Figure 21: Deleted image from MIORe dataset.

Figure 22: Deleted image from VAR-MIORe dataset.

Image Deblurring

Image deblurring restores sharp images from motion, defocus, and other blur by estimating lost detail and modeling image formation.

See all 62 papers

PickMoment: Continuous-Time Single-Image-to-Video via Learning Deblurring and Blur-to-Video
Oct 1, 2026Junseong Shin, Hyeonsu Jo, Daehyun Kim +1Image-to-Video Generation Image Deblurring
Department of Artificial Intelligence, Hanyang University · Department of Intelligence Convergence, Hanyang University · Department of Computer Science, Hanyang University
EvBS: Event-guided Blur Synthesis for Domain-adaptive Motion Deblurring
Aug 8, 2026Junsik Jung, Seokryun Choi, Yoonki Cho +3Image Deblurring Unsupervised Domain Adaptation
Korea Advanced Institute of Science and Technology Daejeon, Republic of Korea
BluRef: Unsupervised Image Deblurring with Dense-Matching References
Mar 15, 2026Bang-Dang Pham, Anh Tran, Cuong Pham +1Image Deblurring
University of Wisconsin-Madison · Qualcomm AI Research · Posts & Telecommunications Inst. of Tech. +1

Oct 1, 2026J/K move · Enter open · S save

PickMoment: Continuous-Time Single-Image-to-Video via Learning Deblurring and Blur-to-Video

Junseong Shin, Hyeonsu Jo, Daehyun Kim, Tae Hyun Kim

Department of Artificial Intelligence, Hanyang University · Department of Intelligence Convergence, Hanyang University · Department of Computer Science, Hanyang University

Image-to-Video Generation Image Deblurring

Open paper AskPDF arXiv

Motion blur arises from the temporal integration of a continuous sharp signal over a finite exposure window, yet existing learning-based methods sidestep this physical model and predict only the sharp signal itself: most single-image deblurring methods recover a single frame at the exposure center, while blur-to-video methods predict a fixed set of frames. We introduce PickMoment, a continuous-time reformulation that directly learns the interval-mean blur over arbitrary sub-intervals of the exposure with a single deterministic model. Drawing an analogy to MeanFlow's average-velocity formulation, we train the model with three supervisions derived from the blur integral: an empirical reconstruction loss from available subframes, an additivity loss that enforces self-consistency across overlapping sub-intervals, and a sharp-frame loss anchored at the zero-interval limit. A single trained model unifies single-image deblurring, blur-to-video generation, and continuous-time pick-a-moment recovery as different queries to the same network, with no separate training for each task. Our PickMoment achieves state-of-the-art performance among generative-based deblurring methods on GoPro and HIDE while competitive against restoration-based methods on RealBlur, and the highest per-frame fidelity on GoPro-7 blur-to-video, all in a single forward pass without iterative sampling.

Abstract

Image Deblurring

Explore similar work

PickMoment: Continuous-Time Single-Image-to-Video via Learning Deblurring and Blur-to-Video

EvBS: Event-guided Blur Synthesis for Domain-adaptive Motion Deblurring

BluRef: Unsupervised Image Deblurring with Dense-Matching References