Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

XYScanNet: A State Space Model for Single Image Deblurring

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A slice-and-scan state-space model for image deblurring beats prior perceptual scores and cuts training time by more than half.

desk verdict Useful Mamba deblurring architecture with big efficiency gains, but the headline perceptual SOTA claim is undercut by the paper's own admission that the loss drives it. read the letter →

arxiv 2412.10338 v3 pith:GT6WWSZH submitted 2024-12-13 cs.CV

Hanzhou Liu , Chengkai Liu , Jiacong Xu , Peng Jiang , Mi Lu This is my paper · ORCID
classification cs.CV
keywords singleimagedeblurringstatespacemodelsMambaslice-and-scanperceptualqualityvisionmodulefeaturefusionrestoration
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that the standard flatten-and-scan strategy used by Mamba-based image restoration models corrupts spatial layout by placing distant pixels next to each other, and that this hurts deblurring sharpness. It proposes a slice-and-scan alternative that scans along image rows and columns, with lightweight inter-slice scans for global blur, and packages this in a new Vision State Space Module. The resulting network, XYScanNet, reports state-of-the-art perceptual quality (KID 0.073, FID 0.138, LPIPS 0.067 on GoPro) while keeping PSNR and SSIM competitive, and it cuts training time by 56.61% and inference memory by 46.18% relative to MambaIR's VSSM in the same framework. A sympathetic reader would care because it offers a concrete architectural reason—spatial misalignment—for why prior Mamba deblurring underperforms, and a cheaper way to fix it.

What carries the argument

The central object is the slice-and-scan strategy inside a Vision State Space Module (VSSM). Instead of flattening a feature map into one long 1D sequence, Intra-Scanner-H slices along the height dimension so each row is scanned separately, and Intra-Scanner-V does the same along width; Inter-Scanners pool the orthogonal dimension (compression factor δ) and use the scanned output as a sigmoid gate over the original features. The VSSM applies layer norm, pixel-wise and depth-wise convolutions, splits channels into vertical/horizontal branches, and combines the dual-scanner output with a SiLU-activated path via element-wise multiplication and a residual addition. The interleaved placement of Intra- and Inter-VSSMs is what lets the network handle both local detail and large-area blur, and the DGFF module uses KL-divergence-gated feature fusion across levels.

What would settle it

Retrain XYScanNet with the same loss but with its Intra/Inter scanners replaced by MambaIR's flatten-and-scan VSSM, or retrain it without the VGG feature-distance term (λ2Lp); if KID, FID, and LPIPS no longer beat the baselines, the perceptual claim would lose its support. The efficiency claim could also be checked by profiling both VSSMs on identical hardware with identical framework settings.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that replacing flatten-and-scan with slice-and-scan removes the spatial misalignment artifact in Mamba-based deblurring. Intra-Scanners preserve local pixel adjacency within each row or column, while Inter-Scanners compress the orthogonal dimension and use a gating signal to capture cross-slice dependencies, so the network estimates both fine local blur and large-area blur without quadrupling scan routes. The paper reports that XYScanNet achieves the best perceptual scores among compared methods on GoPro (KID 0.073, FID 0.138, LPIPS 0.067, NIQE 4.05) with PSNR 33.91 dB and SSIM 0.968, and that in an identical framework its VSSM outperforms MambaIR's VSSM while reducing training time by 56.61% and inference memory by 46.18%. The authors also claim the interleaved Intra/Inter design is validated by ablations and that their dual gating feature fusion module halves the cost of the asymmetric feature fusion baseline.

Load-bearing premise

The load-bearing premise is that the new slice-and-scan module, not the extra VGG-based perceptual loss added during training, is what produces the reported perceptual improvements over methods trained without that loss.

Editorial extensions

If this is right

  • On GoPro, XYScanNet reports the lowest KID (0.073), FID (0.138), and LPIPS (0.067) among the compared methods, with PSNR 33.91 dB and SSIM 0.968.
  • The GoPro-trained model transfers to HIDE and RWBI, improving no-reference metrics (e.g., NIQE reduction of 0.163 on RWBI) and competitive PSNR on HIDE.
  • Replacing half of the Intra-Scanners with Inter-Scanners cuts training time by 20.39% and memory by 9.52% with only a 0.02 dB PSNR drop on GoPro, supporting the interleaved design.
  • In the same baseline framework, the new VSSM beats MambaIR's VSSM on perceptual and many distortion metrics while reducing training time by 56.61% and inference memory by 46.18%.
  • The DGFF fusion module halves FLOPs (by 49.3%) and parameters (by 53.11%) relative to AFF while improving LPIPS on RealBlur-J.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The perceptual gains are partly attributed by the paper itself to the loss function (Charbonnier plus edge plus VGG feature distance), so an ablation that removes the VGG term from XYScanNet's loss would separate the architecture's contribution from the loss's contribution; no such ablation is reported.
  • If slice-and-scan is the active ingredient, the same Intra/Inter scanner design should improve other Mamba-based restoration tasks such as denoising and super-resolution, where flatten-and-scan also introduces spatial misalignment.
  • The inter-slice compression via global average pooling suggests a general recipe for global context in SSMs: scan a cheap pooled projection and use it as a gate, rather than scanning every pixel in four directions.
  • A controlled comparison that trains MambaIR's VSSM and XYScanNet's VSSM with identical losses (including the VGG term) would tell whether the efficiency and quality advantages persist under equal training conditions; the paper only compares with the same framework and training strategy but not with the same loss.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes XYScanNet, a U-Net for single-image deblurring built on a new 'slice-and-scan' state-space module (VSSM) that alternates Intra-Scanners (pixel-level horizontal/vertical scans within image rows/columns) and Inter-Scanners (global-average-pooled slice-level scans) to avoid the spatial misalignment of flatten-and-scan strategies. It also introduces a dual-gating feature fusion (DGFF) module. Experiments on GoPro, HIDE, RWBI, and RealBlur report competitive PSNR/SSIM, improved perceptual metrics (KID/FID/LPIPS/NIQE/Q-ALIGN), and substantially lower training time and memory than MambaIR's VSSM in a matched framework (Table 7). The central claims are a new state-of-the-art in perceptual deblurring quality and a more efficient Mamba restoration block.

Significance. If the architectural claims hold, the slice-and-scan design is a useful contribution to the Mamba-based restoration literature: it provides a plausible mechanism to preserve local spatial structure while capturing global blur, and the matched VSSM comparison in Table 7 is a strong, internally controlled piece of evidence for the efficiency gains (56.61% training-time reduction, 46.18% inference-memory reduction). The paper also releases code and reports a wider range of perceptual metrics than PSNR alone, which is commendable. However, the headline perceptual superiority is not yet attributable to the architecture because the loss function is not controlled across methods and no ablation isolates the perceptual loss; the paper's own caption and limitation paragraph concede this confound. The efficiency and architectural-validity claims are substantially better supported than the perceptual-SOTA claim.

major comments (3)
  1. [Sec. 4.1, Eq. (5); Tables 1–4; Limitation paragraph] The headline claim that XYScanNet achieves state-of-the-art perceptual quality is confounded by the training loss. The loss in Eq. (5) includes a VGG perceptual term Lp with λ2=0.0005, and the Table 1 caption itself states that the performance 'can be attributed to the loss function discussed in Sec. 4.1.' None of the compared methods in Tables 1–4 is retrained with this loss, and no ablation removes Lp from XYScanNet. Consequently the KID/FID/LPIPS improvements could be caused by the objective rather than by the slice-and-scan VSSM. The matched VSSM comparison in Table 7, which holds the framework and training strategy fixed, actually shows KID worsening from 0.121 to 0.129 while FID improves from 0.232 to 0.210, so it does not establish perceptual superiority. The Limitation section's admission that 'further discussions are needed to explore the relationship between loss functions and performance' confirms this gap. Please add an ablation that trains XYScanNet without Lp and, ideally, retrains at least the nearest competitor with the same loss.
  2. [Sec. 4.3, Table 5] The text states that the interleaved intra/inter configuration 'improves PSNR on GoPro by 0.2 dB over the inter-only method,' but Table 5 reports inter-only PSNR 32.18 dB and interleaved 32.11 dB, i.e., a decrease of 0.07 dB. The 0.20 dB improvement is over the intra-only row (31.91 dB), not the inter-only row. The same paragraph's claim of a 20.39% training-time reduction and a 9.52% memory reduction 'compared to the intra-only approach' is also incorrect: those percentages are relative to the inter-only row (103 s to 82 s; 12.82 GB to 11.60 GB), while the intra-only row has lower time and memory (65 s; 10.36 GB). These baseline errors invert the direction of the efficiency/quality trade-off in the ablation and must be corrected.
  3. [Supplementary Sec. 7.1; Tables 1–7] The supplementary states that each network was trained and tested four times to minimize the effects of randomness, yet no table reports variance or significance tests. Several of the reported advantages are tiny (e.g., Table 2 NIQE 3.383 vs. 3.384 for LoFormer; Table 3 LPIPS 0.091 vs. 0.093) and may fall within run-to-run noise. Please report mean ± std (or confidence intervals) for at least the key perceptual metrics on GoPro, HIDE, and RWBI, and clarify how many seeds underlie the numbers in Tables 1–4. Without this, the 'state-of-the-art perceptual quality' claim is not robustly supported.
minor comments (4)
  1. [Sec. 3.2, Inter-Scanner paragraph] The sentence 'The horizontal Intra-Scanner is constructed symmetrically' should read 'Inter-Scanner-H' rather than 'Intra-Scanner'; the surrounding text is describing the symmetric horizontal counterpart of the Inter-Scanner.
  2. [Supplementary Sec. 6.2] The sentence 'As shown in Tab. 7, our slice-and-scan-based VSSM.' is an incomplete fragment; please complete the sentence or remove it.
  3. [Eq. (4) and surrounding text] The KL-divergence terms in Eq. (4) require non-negative, normalized inputs, but X'_cur and X'_oth are outputs of convolutions that may be negative or unnormalized; please specify how the feature maps are converted to probability distributions (e.g., softmax or abs + normalization) and clarify the intended direction of the asymmetric KL term.
  4. [Sec. 4.3, first paragraph; Table 6] The phrase 'LPIPS st by 0.041 on RealBlur-J' contains a typo ('st') and the reported reduction is 0.041 in LPIPS; please correct the wording so the metric and its improvement are unambiguous.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's claims are empirical benchmark measurements with no derivation that reduces to its own inputs.

full rationale

XYScanNet makes architectural and empirical claims rather than deriving predictions from fitted parameters. Equations (1)-(3) define the Intra-/Inter-Scanner operations and the VSSM output in terms of the input features, Mamba-S6, and learned convolutions; no output quantity is defined as the thing it is later claimed to predict. The perceptual SOTA claim rests on Table 1 benchmark numbers, which are external measurements, not a construction. The Table 1 caption states 'This performance can be attributed to the loss function discussed in Sec. 4.1,' and the Limitation section concedes that 'further discussions are needed to explore the relationship between loss functions and performance'; this is an attribution/ablation weakness, not circularity, because the loss function is not a fitted parameter that is renamed as a prediction. The matched-baseline comparison in Table 7 (same framework, training strategy, and similar network size) independently supports the efficiency claims. The few self-citations ([28], [29]) are used for domain adaptation and metric choices and are not load-bearing for the central architecture claim. No uniqueness theorem, ansatz-smuggling citation, or self-definitional reduction is present. The derivation chain is self-contained; concerns about whether the loss function rather than the VSSM causes the perceptual gains are correctness risks, not circularity.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim is an empirical architecture improvement, not a derivation. The main non-standard assumptions are about what row and column slicing and GAP compression preserve; the loss weights are hand-set and directly shape the perceptual metrics. No entirely new conceptual entities are introduced, though the Intra/Inter-Scanner modules are new architectural components.

free parameters (5)
  • Edge loss weight lambda1 = 0.05
    Set manually in Sec. 4.1 Eq. (5); directly influences sharpness and affects distortion metrics.
  • Perceptual (VGG) loss weight lambda2 = 0.0005
    Set manually in Sec. 4.1; this loss is the paper's stated reason for perceptual-metric gains (Table 1 caption). No sensitivity analysis is reported.
  • Inter-Scanner compression factor delta = 1/W, implemented as global average pooling
    Chosen 'for simplicity' in Sec. 3.2 Eq. (2); determines how much cross-slice information is retained and is not ablated.
  • Network width and depth = C=144, blocks [3,3,6]
    Set in Sec. 6.1; capacity choices affect both metric scores and the reported efficiency numbers, with no sensitivity analysis.
  • Training patch size and epoch count = 128x128 patches, 3K epochs for ablations
    Training settings in Sec. 4.1 and Sec. 7.1; these hand-set values affect convergence and final metric values but are standard for the field.
assumptions (4)
  • domain assumption The S6 selective-scan layer, applied to reshaped row and column slices, is a valid sequence model for local and global feature propagation.
    Used as a black box in Eq. (1) and Eq. (2); slicing changes token order and boundary context, but no analysis of boundary effects is provided.
  • domain assumption Motion blur can be decomposed into independent horizontal and vertical components.
    The dual-branch Intra/Inter-Scanners in Sec. 3.2 assume Cartesian decomposition, following prior work [46,48]; diagonal or curved blur may not be captured by this split.
  • ad hoc to paper Global average pooling over the width retains the cross-slice information needed for large-area blur estimation.
    Inter-Scanner Eq. (2) pools the entire width into one value; this architectural choice is unique to this paper and is not compared with other pooling or compression factors.
  • domain assumption Intra-slicing along rows and columns preserves local pixel dependencies and avoids spatial misalignment.
    The core motivation in Sec. 3.2 and Fig. 2; it is asserted qualitatively and supported by ERF images rather than by a quantitative diagnostic.

how reviews work

0 comments
Cite this review

Pith. "Pith review of XYScanNet: A State Space Model for Single Image Deblurring." pith.science (2026). https://pith.science/paper/GT6WWSZH

@misc{pith2026241210338,
  author       = {Pith},
  title        = {Pith review of: XYScanNet: A State Space Model for Single Image Deblurring},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GT6WWSZH}},
  note         = {Machine review of arXiv:2412.10338}
}
abstract

Deep state-space models (SSMs), like recent Mamba architectures, are emerging as a promising alternative to CNN and Transformer networks. Existing Mamba-based restoration methods process visual data by leveraging a flatten-and-scan strategy that converts image patches into a 1D sequence before scanning. However, this scanning paradigm ignores local pixel dependencies and introduces spatial misalignment by positioning distant pixels incorrectly adjacent, which reduces local noise-awareness and degrades image sharpness in low-level vision tasks. To overcome these issues, we propose a novel slice-and-scan strategy that alternates scanning along intra- and inter-slices. We further design a new Vision State Space Module (VSSM) for image deblurring, and tackle the inefficiency challenges of the current Mamba-based vision module. Building upon this, we develop XYScanNet, an SSM architecture integrated with a lightweight feature fusion module for enhanced image deblurring. XYScanNet, maintains competitive distortion metrics and significantly improves perceptual performance. Experimental results show that XYScanNet enhances KID by $17\%$ compared to the nearest competitor.

Figures

Figures reproduced from arXiv: 2412.10338 by the authors.

Figure 1
Figure 1. XYScanNet achieves state-of-the-art performance on [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. (a) The two issues of flatten-and-scan strategies in a single scanning route: adjacent [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. XYScanNet is a three-level U-Net. (a) An encoder block. (b) A decoder block. (c) Intra-VSSM incorporates 3 × 3 depth-wise convolutions to prevent local pixel forgetting [17] and the dual-branch Intra-Scanners (refer to [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (16 more)
Figure 4
Figure 4. Figure 4: Structures of (a) asymmetric feature fusion (AFF) [ [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: From the top to bottom rows: qualitative comparisons on the GoPro [ [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Replacing half of the Intra-Scanners with Inter-Scanners (refer to Fig. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: On real-world RWBI [57], our DGFF is able to estimate the blur patterns that are ignored by skip connections and AFF [6]. AFF DGFF NIQE↓ LPIPS↓ Flops Params HIDE RWBI HIDE RB.J (G) (K) 3.46 3.52 0.125 0.889 - - ✓ 3.40 3.53 0.124 0.878 4.32 152.35 ✓ 3.41 3.51 0.122 0.84…
Figure 8
Figure 8. Figure 8: Qualitative comparisons of the proposed VSSM versus [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 10
Figure 10. Figure 10: The hybrid method captures global and uniform blurred [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Supplementary to Fig [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Qualitative results of deblurring models trained and tested on the RealBlur-R dataset [ [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]
Figure 13
Figure 13. Figure 13: Qualitative results of deblurring models trained and tested on RealBlur-R [ [PITH_FULL_IMAGE:figures/full_fig_p014_13.png]
Figure 14
Figure 14. Figure 14: Qualitative results of deblurring models trained and tested on RealBlur-J [ [PITH_FULL_IMAGE:figures/full_fig_p015_14.png]
Figure 15
Figure 15. Figure 15: Qualitative results of deblurring models trained and tested on RealBlur-J [ [PITH_FULL_IMAGE:figures/full_fig_p016_15.png]
Figure 16
Figure 16. Figure 16: Qualitative results of deblurring models trained and tested on RealBlur-J [ [PITH_FULL_IMAGE:figures/full_fig_p017_16.png]
Figure 17
Figure 17. Figure 17: Qualitative results of deblurring models trained and tested on RealBlur-J [ [PITH_FULL_IMAGE:figures/full_fig_p018_17.png]
Figure 18
Figure 18. Figure 18: Qualitative results of GoPro-trained deblurring models tested on GoPro [ [PITH_FULL_IMAGE:figures/full_fig_p019_18.png]
Figure 19
Figure 19. Figure 19: Qualitative results of GoPro-trained deblurring models tested on HIDE [ [PITH_FULL_IMAGE:figures/full_fig_p020_19.png]
Figure 20
Figure 20. Figure 20: Qualitative results of GoPro-trained deblurring models tested on RWBI [ [PITH_FULL_IMAGE:figures/full_fig_p021_20.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The Computational Limits of State-Space Models and Mamba via the Lens of Circuit Complexity

    cs.CC 2024-12 conditional novelty 5.0 of 10

    Mamba and state-space models with constant depth and poly(n) precision are shown to lie in DLOGTIME-uniform TC0, the same circuit class as Transformers, so their expressiveness is bounded unless TC0=NC1.

Reference graph

Works this paper leans on

67 extracted references · 45 canonical work pages · cited by 1 Pith paper

  1. [1]

    Demystifying mmd gans

    Mikołaj Bi ´nkowski, Danica J Sutherland, Michael Arbel, and Arthur Gretton. Demystifying mmd gans. arXiv preprint arXiv:1801.01401, 2018. 2

    arXiv 2018

  2. [2]

    The perception-distortion tradeoff

    Yochai Blau and Tomer Michaeli. The perception-distortion tradeoff. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 6228–6237, 2018. 5

    work page 2018

  3. [3]

    Rethinking lossy com- pression: The rate-distortion-perception tradeoff

    Yochai Blau and Tomer Michaeli. Rethinking lossy com- pression: The rate-distortion-perception tradeoff. In Inter- national Conference on Machine Learning , pages 675–685. PMLR, 2019. 5

    work page 2019

  4. [4]

    Blind image deblurring with local maximum gra- dient prior

    Liang Chen, Faming Fang, Tingting Wang, and Guixu Zhang. Blind image deblurring with local maximum gra- dient prior. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1742–1750,

    work page

  5. [5]

    Simple baselines for image restoration

    Liangyu Chen, Xiaojie Chu, Xiangyu Zhang, and Jian Sun. Simple baselines for image restoration. In European confer- ence on computer vision, pages 17–33. Springer, 2022. 5, 6, 8, 9, 10

    work page 2022

  6. [6]

    Rethinking coarse-to-fine approach in sin- gle image deblurring

    Sung-Jin Cho, Seo-Won Ji, Jun-Pyo Hong, Seung-Won Jung, and Sung-Jea Ko. Rethinking coarse-to-fine approach in sin- gle image deblurring. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 4641–4650,

  7. [7]

    Sigmoid- weighted linear units for neural network function approxima- tion in reinforcement learning

    Stefan Elfwing, Eiji Uchibe, and Kenji Doya. Sigmoid- weighted linear units for neural network function approxima- tion in reinforcement learning. Neural networks, 107:3–11,

  8. [8]

    Self-supervised non-uniform kernel estimation with flow-based motion prior for blind im- age deblurring

    Zhenxuan Fang, Fangfang Wu, Weisheng Dong, Xin Li, Jin- jian Wu, and Guangming Shi. Self-supervised non-uniform kernel estimation with flow-based motion prior for blind im- age deblurring. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 18105– 18114, 2023. 5, 6, 2, 3, 4, 7, 8, 9, 10

    work page 2023

Show all 67 references
  1. [9]

    Removing camera shake from a single photograph

    Rob Fergus, Barun Singh, Aaron Hertzmann, Sam T Roweis, and William T Freeman. Removing camera shake from a single photograph. In ACM transactions on graphics (TOG), pages 787–794. ACM, 2006. 1, 2

    2006

  2. [10]

    Hungry hungry hippos: To- wards language modeling with state space models

    Daniel Y Fu, Tri Dao, Khaled K Saab, Armin W Thomas, Atri Rudra, and Christopher R´e. Hungry hungry hippos: To- wards language modeling with state space models. arXiv preprint arXiv:2212.14052, 2022. 2

    2022 arXiv

  3. [11]

    Learning enriched features via selective state spaces model for efficient image deblurring

    Hu Gao, Bowen Ma, Ying Zhang, Jingfan Yang, Jing Yang, and Depeng Dang. Learning enriched features via selective state spaces model for efficient image deblurring. In Pro- ceedings of the 32nd ACM International Conference on Mul- timedia, pages 710–718, 2024. 2, 3, 5, 6

    2024

  4. [12]

    Shift-tolerant perceptual similarity metric

    Abhijay Ghildyal and Feng Liu. Shift-tolerant perceptual similarity metric. In European Conference on Computer Vi- sion, pages 91–107. Springer, 2022. 2

    2022

  5. [13]

    Mamba: Linear-time sequence modeling with selective state spaces

    Albert Gu and Tri Dao. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752, 2023. 1, 2, 3, 4

    2023 arXiv

  6. [14]

    Hippo: Recurrent memory with optimal polynomial projections

    Albert Gu, Tri Dao, Stefano Ermon, Atri Rudra, and Christo- pher R´e. Hippo: Recurrent memory with optimal polynomial projections. Advances in neural information processing sys- tems, 33:1474–1487, 2020

    2020

  7. [15]

    Efficiently modeling long sequences with structured state spaces

    Albert Gu, Karan Goel, and Christopher R ´e. Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396, 2021. 3

    2021 arXiv

  8. [16]

    Combining recurrent, convolutional, and continuous-time models with linear state space layers

    Albert Gu, Isys Johnson, Karan Goel, Khaled Saab, Tri Dao, Atri Rudra, and Christopher R ´e. Combining recurrent, convolutional, and continuous-time models with linear state space layers. Advances in neural information processing sys- tems, 34:572–585, 2021. 2

    2021

  9. [17]

    Mambair: A simple baseline for image restoration with state-space model

    Hang Guo, Jinmin Li, Tao Dai, Zhihao Ouyang, Xudong Ren, and Shu-Tao Xia. Mambair: A simple baseline for image restoration with state-space model. arXiv preprint arXiv:2402.15648, 2024. 1, 2, 3, 8

    2024 arXiv

  10. [18]

    Gaussian error linear units (gelus)

    Dan Hendrycks and Kevin Gimpel. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415, 2016. 4

    2016 arXiv

  11. [19]

    Gans trained by a two time-scale update rule converge to a local nash equilib- rium

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilib- rium. Advances in neural information processing systems , 30, 2017. 2

    2017

  12. [20]

    Image deblurring and denoising using color priors

    Neel Joshi, C Lawrence Zitnick, Richard Szeliski, and David J Kriegman. Image deblurring and denoising using color priors. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, pages 1550–1557. IEEE, 2009. 2

    2009

  13. [21]

    Single-image deblurring with neural networks: A comparative survey

    Jaihyun Koh, Jangho Lee, and Sungroh Yoon. Single-image deblurring with neural networks: A comparative survey. Computer Vision and Image Understanding , 203:103134,

  14. [22]

    Efficient frequency domain-based trans- formers for high-quality image deblurring

    Lingshun Kong, Jiangxin Dong, Jianjun Ge, Mingqiang Li, and Jinshan Pan. Efficient frequency domain-based trans- formers for high-quality image deblurring. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 5886–5895, 2023. 1, 2, 5, 6...

    2023

  15. [23]

    Deblurgan: Blind mo- tion deblurring using conditional adversarial networks

    Orest Kupyn, V olodymyr Budzan, Mykola Mykhailych, Dmytro Mishkin, and Ji ˇr´ı Matas. Deblurgan: Blind mo- tion deblurring using conditional adversarial networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8183–8192, 2018. 1

    2018

  16. [24]

    Deblurgan-v2: Deblurring (orders-of-magnitude) faster and better

    Orest Kupyn, Tetiana Martyniuk, Junru Wu, and Zhangyang Wang. Deblurgan-v2: Deblurring (orders-of-magnitude) faster and better. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 8878–8887,

  17. [25]

    Advanced multi- ple linear regression based dark channel prior applied on de- hazing image and generating synthetic haze

    Binghan Li, Yindong Hua, and Mi Lu. Advanced multi- ple linear regression based dark channel prior applied on de- hazing image and generating synthetic haze. arXiv preprint arXiv:2103.07065, 2021. 2

    2021 arXiv

  18. [26]

    Perceptual variousness motion deblurring with light global context refinement

    Jichun Li, Weimin Tan, and Bo Yan. Perceptual variousness motion deblurring with light global context refinement. In Proceedings of the IEEE/CVF international conference on computer vision, pages 4116–4125, 2021. 1

    2021

  19. [27]

    Motion-adaptive separable collaborative filters for blind motion deblurring

    Chengxu Liu, Xuan Wang, Xiangyu Xu, Ruhao Tian, Shuai Li, Xueming Qian, and Ming-Hsuan Yang. Motion-adaptive separable collaborative filters for blind motion deblurring. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 25595–25605, ...

    2024

  20. [28]

    Real- world image deblurring via unsupervised domain adaptation

    Hanzhou Liu, Binghan Li, Mi Lu, and Yucheng Wu. Real- world image deblurring via unsupervised domain adaptation. In International Symposium on Visual Computing , pages 148–159, 2023. 1, 2

    2023

  21. [29]

    De- blurdinat: A lightweight and effective transformer for image deblurring

    Hanzhou Liu, Binghan Li, Chengkai Liu, and Mi Lu. De- blurdinat: A lightweight and effective transformer for image deblurring. arXiv preprint arXiv:2403.13163, 2024. 2, 4, 1

    2024 arXiv

  22. [30]

    Vmamba: Visual state space model

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, and Yunfan Liu. Vmamba: Visual state space model. arXiv preprint arXiv:2401.10166,

    arXiv

  23. [31]

    Intriguing findings of frequency selection for image deblurring

    Xintian Mao, Yiming Liu, Fengze Liu, Qingli Li, Wei Shen, and Yan Wang. Intriguing findings of frequency selection for image deblurring. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 1905–1913, 2023. 5, 6, 2, 3, 4, 7

    1905

  24. [32]

    Loformer: Local frequency transformer for im- age deblurring

    Xintian Mao, Jiansheng Wang, Xingran Xie, Qingli Li, and Yan Wang. Loformer: Local frequency transformer for im- age deblurring. In Proceedings of the 32nd ACM Inter- national Conference on Multimedia , pages 10382–10391,

  25. [33]

    Long range language modeling via gated state spaces

    Harsh Mehta, Ankit Gupta, Ashok Cutkosky, and Behnam Neyshabur. Long range language modeling via gated state spaces. arXiv preprint arXiv:2206.13947, 2022. 2

    2022 arXiv

  26. [34]

    completely blind

    Anish Mittal, Rajiv Soundararajan, and Alan C Bovik. Mak- ing a “completely blind” image quality analyzer. IEEE Sig- nal processing letters, 20(3):209–212, 2012. 2

    2012

  27. [35]

    Deep multi-scale convolutional neural network for dynamic scene deblurring

    Seungjun Nah, Tae Hyun Kim, and Kyoung Mu Lee. Deep multi-scale convolutional neural network for dynamic scene deblurring. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3883–3891,

  28. [36]

    Blind image deblurring using dark channel prior

    Jinshan Pan, Deqing Sun, Hanspeter Pfister, and Ming- Hsuan Yang. Blind image deblurring using dark channel prior. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1628–1636, 2016. 1, 2

    2016

  29. [37]

    On aliased resizing and surprising subtleties in gan evaluation

    Gaurav Parmar, Richard Zhang, and Jun-Yan Zhu. On aliased resizing and surprising subtleties in gan evaluation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11410–11420, 2022. 5

    2022

  30. [38]

    Mamba-360: Survey of state space models as transformer alternative for long sequence modelling: Methods, appli- cations, and challenges

    Badri Narayana Patro and Vijay Srinivas Agneeswaran. Mamba-360: Survey of state space models as transformer alternative for long sequence modelling: Methods, appli- cations, and challenges. arXiv preprint arXiv:2404.16112 ,

    arXiv

  31. [39]

    Hyena hierarchy: Towards larger con- volutional language models

    Michael Poli, Stefano Massaroli, Eric Nguyen, Daniel Y Fu, Tri Dao, Stephen Baccus, Yoshua Bengio, Stefano Ermon, and Christopher R ´e. Hyena hierarchy: Towards larger con- volutional language models. In International Conference on Machine Learning, pages 28043–28078. PMLR, 2023. 2

    2023

  32. [40]

    Prajit Ramachandran, Barret Zoph, and Quoc V . Le. Swish: a self-gated activation function. arXiv: Neural and Evolu- tionary Computing, 2017. 4

    2017

  33. [41]

    Image deblurring via en- hanced low-rank prior

    Wenqi Ren, Xiaochun Cao, Jinshan Pan, Xiaojie Guo, Wang- meng Zuo, and Ming-Hsuan Yang. Image deblurring via en- hanced low-rank prior. IEEE Transactions on Image Pro- cessing, 25(7):3426–3437, 2016. 2

    2016

  34. [42]

    Real-world blur dataset for learning and benchmarking de- blurring algorithms

    Jaesung Rim, Haeyun Lee, Jucheol Won, and Sunghyun Cho. Real-world blur dataset for learning and benchmarking de- blurring algorithms. In Proceedings of the European Con- ference on Computer Vision (ECCV) , 2020. 5, 6, 7, 2, 3, 4

    2020

  35. [43]

    Human-aware motion deblurring

    Ziyi Shen, Wenguan Wang, Xiankai Lu, Jianbing Shen, Haibin Ling, Tingfa Xu, and Ling Shao. Human-aware motion deblurring. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 5572–5581,

  36. [44]

    Simplified state space layers for sequence modeling

    Jimmy TH Smith, Andrew Warrington, and Scott W Linder- man. Simplified state space layers for sequence modeling. arXiv preprint arXiv:2208.04933, 2022. 2

    2022 arXiv

  37. [45]

    Spatially-attentive patch-hierarchical network for adaptive motion deblurring

    Maitreya Suin, Kuldeep Purohit, and AN Rajagopalan. Spatially-attentive patch-hierarchical network for adaptive motion deblurring. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 3606–3615, 2020. 1

    2020

  38. [46]

    Learn- ing a convolutional neural network for non-uniform motion blur removal

    Jian Sun, Wenfei Cao, Zongben Xu, and Jean Ponce. Learn- ing a convolutional neural network for non-uniform motion blur removal. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 769–777,

  39. [47]

    Scale-recurrent network for deep image deblurring

    Xin Tao, Hongyun Gao, Xiaoyong Shen, Jue Wang, and Ji- aya Jia. Scale-recurrent network for deep image deblurring. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8174–8182, 2018. 2

    2018

  40. [48]

    Stripformer: Strip transformer for fast image deblurring

    Fu-Jen Tsai, Yan-Tsung Peng, Yen-Yu Lin, Chung-Chi Tsai, and Chia-Wen Lin. Stripformer: Strip transformer for fast image deblurring. In European conference on computer vi- sion, pages 146–162. Springer, 2022. 1, 2, 3, 4

    2022

  41. [49]

    Uformer: A general u-shaped transformer for image restoration

    Zhendong Wang, Xiaodong Cun, Jianmin Bao, Wengang Zhou, Jianzhuang Liu, and Houqiang Li. Uformer: A general u-shaped transformer for image restoration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 17683–17693, 2022. 1, 2

    2022

  42. [50]

    De- blurring via stochastic refinement

    Jay Whang, Mauricio Delbracio, Hossein Talebi, Chitwan Saharia, Alexandros G Dimakis, and Peyman Milanfar. De- blurring via stochastic refinement. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16293–16303, 2022. 2, 5, 6

    2022

  43. [51]

    Q-align: Teaching lmms for visual scoring via discrete text-defined levels

    Haoning Wu, Zicheng Zhang, Weixia Zhang, Chaofeng Chen, Liang Liao, Chunyi Li, Yixuan Gao, Annan Wang, Erli Zhang, Wenxiu Sun, et al. Q-align: Teaching lmms for visual scoring via discrete text-defined levels. arXiv preprint arXiv:2312.17090, 2023. 6, 2

    2023 arXiv

  44. [52]

    Image deblurring via extreme channels prior

    Yanyang Yan, Wenqi Ren, Yuanfang Guo, Rui Wang, and Xiaochun Cao. Image deblurring via extreme channels prior. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4003–4011, 2017. 1, 2

    2017

  45. [53]

    Multi-stage progressive image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, Ming-Hsuan Yang, and Ling Shao. Multi-stage progressive image restoration. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14821–14831, 2021. 1, 2, 5, ...

    2021

  46. [54]

    Restormer: Efficient transformer for high-resolution image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Mu- nawar Hayat, Fahad Shahbaz Khan, and Ming-Hsuan Yang. Restormer: Efficient transformer for high-resolution image restoration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5728–5739,

  47. [55]

    Multi- image blind deblurring using a coupled adaptive sparse prior

    Haichao Zhang, David Wipf, and Yanning Zhang. Multi- image blind deblurring using a coupled adaptive sparse prior. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1051–1058, 2013. 2

    2013

  48. [56]

    Deep stacked hierarchical multi-patch network for image deblurring

    Hongguang Zhang, Yuchao Dai, Hongdong Li, and Piotr Ko- niusz. Deep stacked hierarchical multi-patch network for image deblurring. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 5978–5986,

  49. [57]

    Deblurring by realistic blurring

    Kaihao Zhang, Wenhan Luo, Yiran Zhong, Lin Ma, Bjorn Stenger, Wei Liu, and Hongdong Li. Deblurring by realistic blurring. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2737–2746,

  50. [58]

    Deep image deblurring: A survey

    Kaihao Zhang, Wenqi Ren, Wenhan Luo, Wei-Sheng Lai, Bj¨orn Stenger, Ming-Hsuan Yang, and Hongdong Li. Deep image deblurring: A survey. International Journal of Com- puter Vision, 130(9):2103–2130, 2022. 1, 2

    2022

  51. [59]

    Perception- distortion balanced admm optimization for single-image super-resolution

    Yuehan Zhang, Bo Ji, Jia Hao, and Angela Yao. Perception- distortion balanced admm optimization for single-image super-resolution. In European Conference on Computer Vi- sion, pages 108–125. Springer, 2022. 5

    2022

  52. [60]

    Vision mamba: Efficient visual representation learning with bidirectional state space model

    Lianghui Zhu, Bencheng Liao, Qian Zhang, Xinlong Wang, Wenyu Liu, and Xinggang Wang. Vision mamba: Efficient visual representation learning with bidirectional state space model. arXiv preprint arXiv:2401.09417, 2024. 1, 2

    2024 arXiv

  53. [61]

    Perceptual-distortion balanced image super-resolution is a multi-objective optimization problem

    Qiwen Zhu, Yanjie Wang, Shilv Cai, Liqun Chen, Ji- ahuan Zhou, Luxin Yan, Sheng Zhong, and Xu Zou. Perceptual-distortion balanced image super-resolution is a multi-objective optimization problem. In Proceedings of the 32nd ACM International Conference on Multimedia , pages 748...

    2024

  54. [62]

    Barriers towards no-reference met- rics application to compressed video quality analysis: On the example of no-reference metric niqe

    Anastasia Zvezdakova, Dmitriy Kulikov, Denis Kondranin, and Dmitriy Vatolin. Barriers towards no-reference met- rics application to compressed video quality analysis: On the example of no-reference metric niqe. arXiv preprint arXiv:1907.03842, 2019. 6, 2 XYScanNet: A State Spa...

    1907 arXiv

  55. [65]

    After that, we explain the basic build- ing blocks with ablation experiments

    Interpretability In this section, we outline the overall architecture of XYScanNet at first. After that, we explain the basic build- ing blocks with ablation experiments. 6.1. Overall Pipeline As shown in Fig. 3, XYScanNet is of an asymmetric U- Net structure with cross-level ...

  56. [66]

    Experimental Settings We train the full XYScanNet on 8 A100 GPUs, each with 40 GB memory

    Experiment Details 7.1. Experimental Settings We train the full XYScanNet on 8 A100 GPUs, each with 40 GB memory. To computer the metric scores, we use a local RTX 3090 GPU. Specifically, we calculate PSNR and SSIM with Matlab functions, while the other metrics by Python. Abla...

  57. [67]

    The low-light images may not be clearly visible in this material

    Performance Analysis In this section, we begin with visual results of deblurring models trained and tested on the RealBlur-R dataset [42], which have not been included in the main paper due to space limits. The low-light images may not be clearly visible in this material. Then...

  58. [2022]

    1, 2, 4, 5, 6, 8, 9, 10

  59. [2024]

    1, 5, 6, 2, 3, 4, 7, 8, 9, 10

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.