Pith. sign in

REVIEW 4 major objections 4 minor 49 references

MPSI: Mamba enhancement model for pixel-wise sequential interaction Image Super-Resolution

T0 review · 4 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 MPSI uses bidirectional Mamba state-space modeling to beat prior lightweight super-resolution models at 3x and 4x upscaling.

desk verdict The Mamba blocks are plausibly designed and the ablations are decent, but the state-of-the-art claim is unsupported by the lightweight non-retrained baselines and is directly contradicted by the x2 Manga109 result. read the letter →

arxiv 2412.07222 v1 pith:MNJCQIB5 submitted 2024-12-10 cs.CV cs.AIeess.IV

Yuchun He , Yuhan He This is my paper · ORCID
classification cs.CVcs.AIeess.IV
keywords singleimagesuper-resolutionMambastatespacemodelpixel-wisesequentialinteractionlightweightchannelattentionrestoration
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

Single-image super-resolution models built on convolutional or windowed-Transformer backbones struggle to connect distant pixels, so the paper introduces MPSI, a network that treats image features as sequences and models their long-range interactions with Mamba-style state-space blocks. MPSI combines spatial-window self-attention with a Channel-Mamba Block that runs features forward and backward through two Mamba layers, plus a Mamba channel recursion module that pools features from every block and reweights the group output by channel. In experiments with bicubic degradation, MPSI reports the best PSNR and SSIM among the compared lightweight methods on all five benchmarks at $\times 3$ and $\times 4$ upscaling, and strong but not uniformly best results at $\times 2$. If those numbers hold under matched training conditions, the paper would show that selective state-space sequence models are a practical backbone for lightweight super-resolution.

What carries the argument

The Channel-Mamba Block (CMB) is the central new object: it uses the Dual Direction Bi-Mamba module (DDBM) to run the feature sequence in both directions through two Mamba state-space layers, governed by $h_t = A h_{t-1} + B x_t$ and $y_t = C h_t$, then sums the forward and backward outputs through a linear projection. Mamba, a selective state-space model, gives linear-time dependency across arbitrarily long feature sequences without position embeddings, which is why the paper expects it to fix the long-range modeling gap of windowed self-attention. The second mechanism, the Mamba channel recursion module (MCRM), applies adaptive average pooling to the input and each block's output in a group, runs those pooled vectors through Mamba, and uses a sigmoid-gated multilayer perceptron to produce a channel-weight vector that multiplies the group's final feature map. MCRM is what lets early-layer features influence the final result without dense skip connections across all layers.

What would settle it

Retrain lightweight SwinIR, ELAN, DAT, and MambaIR under exactly MPSI's training protocol—500K iterations, L1 loss, Adam with $\beta_1=0.9$, $\beta_2=0.99$, learning rate $2\times10^{-4}$ with the stated halving schedule, and the same random rotation/flip augmentation—and compare on Urban100 at $\times 4$. If any of them reaches or exceeds MPSI's reported 32.58 dB PSNR, the paper's state-of-the-art claim would fail; a second check would replace DDBM's reverse Mamba path with an additional forward pass to see whether bidirectionality itself contributes.

Watch

Extended reading notes

Core claim

The paper's central claim is that MPSI notably surpasses existing methods and attains state-of-the-art performance, with the reported evidence concentrated in the $\times 3$ and $\times 4$ upscaling tasks, where it achieves the highest PSNR and SSIM on Set5, Set14, B100, Urban100, and Manga109. Against EDSR, IMDN, LAPAR-A, RDN, ESRT, SwinIR, MambaIR, ELAN, and DAT, the largest recorded margin is on Urban100 at $\times 4$, where MPSI beats the second-best result by 0.07 dB in PSNR and 0.0024 in SSIM. The paper attributes the gain to two mechanisms: bidirectional Mamba sequence modeling in the Channel-Mamba Block, which captures global pixel interactions that windowed attention misses, and the Mamba channel recursion module, which preserves and reweights information from earlier layers. It also notes that the SwinIR, ELAN, DAT, and MambaIR comparison numbers come from lightweight versions of those models.

Load-bearing premise

The state-of-the-art claim rests on the comparability assumption that published PSNR/SSIM numbers for the lightweight baselines can be compared directly with MPSI's single training run, even though no baseline was retrained under MPSI's optimizer, loss, schedule, and augmentation.

Editorial extensions

If this is right

  • MPSI's reported results imply that selective state-space models can replace global self-attention as the long-range modeling component in a lightweight super-resolution network.
  • The pairing of windowed spatial self-attention (STB) with channel-wise bidirectional Mamba (CMB) indicates that spatial and channel interactions can be aggregated without full-image attention, which is cheaper at high resolution.
  • The MCRM design implies that recursive channel weighting can recover some of the benefit of dense feature reuse while keeping the network graph simpler.
  • The Urban100 margin at $\times 4$ suggests MPSI is especially effective on images with repeated structured texture, where long-range pixel dependencies carry real information.
  • The $\times 2$ results, where MPSI is not uniformly best, imply the advantage of the Mamba-based design is more pronounced at larger upscaling factors.

Reading between the lines

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

  • One consequence the paper leaves implicit is that the state-of-the-art claim is tied to lightweight baselines whose published numbers were not reproduced under MPSI's training protocol; retraining full-size baselines under identical conditions could narrow or reverse the reported margins.
  • A testable extension would replace DDBM's reverse Mamba pass with an additional forward pass of the same cost; the paper's ablation replaces DDBM with channel attention, not with a unidirectional Mamba, so the value of bidirectionality itself is not isolated.
  • MCRM's recursive gating could be grafted onto other super-resolution backbones, such as SwinIR or DAT, as a plug-in module; the paper only tests it inside MPSI, so its transferability is unknown.
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

4 major / 4 minor

Summary. The paper proposes MPSI, a single-image super-resolution network built on the Mamba state-space model. The architecture combines Spatial Transformer Blocks and Channel-Mamba Blocks into Spatial Attention Mamba Groups, plus a Mamba-based channel recursion module (MCRM) intended to propagate early-layer features. The authors report experiments on Set5, Set14, B100, Urban100, and Manga109 for ×2, ×3, and ×4 upscaling, comparing against published results of several SR models, and claim state-of-the-art performance. Ablation studies on Urban100 and Manga109 assess the contributions of CMB, MCRM, DDBM, and the Mamba recursive process.

Significance. If the central claim were fully supported, MPSI would be a meaningful contribution to lightweight SR, combining Mamba's long-sequence modeling with a novel feature-recursion mechanism. The paper's strengths include a clearly described architecture, a reasonable set of ablations, and an explicit comparison with several recent Transformer-based SR models. However, the state-of-the-art assertion is not yet secure: the comparison protocol restricts baselines to lightweight versions without retraining them, no error bars or parameter counts are provided, and the reported gains over DAT are small enough to be within run-to-run variation. The inconsistency between the loss stated in Section 3.1 and the loss used in Section 4.1 further undermines reproducibility. The core ideas have merit, but the evidence as presented does not justify the unqualified SOTA claim.

major comments (4)
  1. [Section 4.2 / Table 1] The claim that MPSI 'notably surpasses the existing methods, attaining state-of-the-art performance' (Section 1, Contribution 3) is not supported by the comparison as conducted. The table caption states that SwinIR, ELAN, MambaIR, and DAT results are from lightweight versions and are taken from their papers; none of these baselines is retrained under MPSI's training protocol (batch size 8, patch 64×64, 500K iterations, L1 loss, specific LR schedule). Differences in loss, iteration count, patch size, augmentation, and schedule are uncontrolled confounds. To substantiate a SOTA claim, the authors should either retrain the baselines under the same protocol or report parameter counts, FLOPs, and multiple-seed statistics with significance testing, and restrict the SOTA claim to the controlled comparison.
  2. [Section 3.1 vs. Section 4.1] The loss function is described inconsistently. Section 3.1 states 'Our model is optimized by MSE loss and perceptual loss,' while Section 4.1 states that the Adam optimizer is used 'to minimize the L1 loss function.' These are different objectives, and the discrepancy makes reproduction ambiguous and weakens the comparability with baselines trained with their own published losses. The authors must state exactly which loss (or combination) is used and correct the inconsistent passage.
  3. [Section 4.2 / Table 1] The empirical support for the SOTA claim is internally inconsistent. At ×2, MPSI trails DAT on Manga109 in both PSNR (39.47 vs. 39.49) and SSIM (0.9787 vs. 0.9788), contradicting the abstract's unqualified statement of state-of-the-art performance. In addition, the reported margins over the second-best method are very small (e.g., ×4 Urban100 PSNR +0.07 dB, SSIM +0.0024), and no error bars or multiple-seed results are reported, so these differences may not be significant. The claims in the abstract and conclusion should be qualified to reflect the actual pattern of results.
  4. [Section 4.1 / Table 1] No parameter counts or computational cost are reported for MPSI or the comparison models. Since the paper emphasizes designing a lightweight model and compares only with lightweight versions of baselines, the absence of parameter/FLOP tables makes it impossible to judge whether MPSI's gains are obtained within a comparable parameter budget. The definition of 'lightweight' in Section 2.2 is therefore unsubstantiated, and the comparison may not be parameter-fair.
minor comments (4)
  1. [Table 3] There is a typo in Table 3: 'SIMM' should be 'SSIM'. This appears in the table body and in the surrounding text ('both PSNR and SIMM dropped').
  2. [Section 3.2] The text says 'The way SW-SA handles features in SAT is very similar to DAT [6].' The acronym 'SAT' is not defined; this should likely be 'STB' (Spatial Transformer Block).
  3. [Table 1 / References] The baseline 'MambaIR' is included in Table 1 without a citation in the reference list. A citation to the MambaIR paper should be added.
  4. [Section 4.2] The list of compared models in the text mentions CARN, EDSR, IMDN, LAPAR-A, RDN, ESRT, SwinIR, ELAN, and DAT, but Table 1 also includes MambaIR, and CARN does not appear in Table 1. The model list should be reconciled with the table.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: MPSI is an empirical architecture paper whose performance claims rest on held-out benchmark evaluations, not on fitted inputs or self-citation chains.

full rationale

This paper proposes a new super-resolution architecture (MPSI) built from Mamba-based blocks (Channel-Mamba Block with Dual Direction Bi-Mamba) and a Mamba Channel Recursion Module, then reports PSNR/SSIM on Set5, Set14, B100, Urban100, and Manga109. No predicted constant is fitted and then renamed as a prediction: the network weights are learned on DIV2K/Flickr2K and evaluated on held-out test sets, and the ablations compare trained variants on the same held-out sets. The Section 1 Contribution 3 claim and the Section 4.2 state-of-the-art claim are empirical benchmarking statements, not quantities forced by construction or by a self-citation chain. The paper makes no appeal to a uniqueness theorem from the authors' prior work, and its citations to Mamba, DAT, SwinIR, SENet, and related work are external architectural influences rather than load-bearing self-citations. The manuscript does contain explicit limitations and inconsistencies that affect soundness rather than circularity: Section 4.2 acknowledges that only lightweight versions of SwinIR, ELAN, and DAT were compared, with Table 1 noting that the comparison baselines are taken from their papers; Section 4.2 also concedes that at x2 on Manga109 MPSI trails DAT; and the loss description is inconsistent (Section 3.1 says MSE and perceptual loss, while Section 4.1 says L1 loss). These are reproducibility and comparability concerns, not evidence that any result reduces to its own inputs. Accordingly, no circular step can be exhibited and the score is 0.

Assumptions & free parameters 10 free parameters · 6 assumptions · 0 invented entities

The central claim rests on hand-chosen architecture hyperparameters and on untested empirical assumptions about benchmark comparability and degradation model. No physical or conceptual entities are postulated; the new modules are architectural compositions, not independent entities.

free parameters (10)
  • Number of SAMGs and SAMBs per group = 1 SAMG, 9 SAMBs
    Model depth chosen by hand; directly affects capacity and results.
  • Window size for spatial window self-attention = 8x32
    Manual choice for spatial attention windows; affects receptive field and locality.
  • SGFN channel dimension = 60
    Hand-set feed-forward width with no reported sensitivity analysis.
  • SGFN expansion factor = 2
    Manual expansion ratio in the feed-forward network.
  • DDBM SSM state expansion factor = 32
    Mamba block capacity in CMB; chosen by hand.
  • DDBM local convolution width and block expansion factor = 3 and 4
    Hand-set kernel width and expansion ratio for the Mamba blocks in CMB.
  • MCRM SSM state expansion factor = 64
    Mamba capacity in the recursion module; chosen by hand.
  • MCRM local convolution width and block expansion factor = 4 and 2
    Hand-set kernel width and expansion ratio for MCRM.
  • Number of attention heads = 6
    Hand-set attention configuration in STB.
  • Training iterations and learning rate schedule = 500K iterations; 2e-4 with halving at 250K, 400K, 450K, 475K
    Training recipe for MPSI; baselines use different schedules, making direct comparison an assumption.
assumptions (6)
  • domain assumption Bicubic degradation defines the LR/HR pairs for training and testing.
    Section 4.1 uses bicubic degradation; results may not transfer to real or other degradations.
  • domain assumption PSNR and SSIM on five benchmarks are sufficient to rank SR methods.
    Section 4.1 uses only these metrics, with no perceptual or downstream task evaluation.
  • domain assumption Published baseline numbers are directly comparable without retraining under MPSI's protocol.
    Section 4.2 takes baseline results from their papers; training schedules and loss functions differ.
  • ad hoc to paper Comparing lightweight versions of competing transformers suffices for a SOTA claim.
    Section 4.2 restricts comparisons to lightweight architectures; this premise is untested against full-size models.
  • standard math Mamba's SSM equations apply to flattened image feature sequences.
    Section 3.2 adopts Mamba [12] as-is, with no image-specific adaptation or proof of suitability.
  • ad hoc to paper MCRM's recursive pooling preserves early-layer information useful for reconstruction.
    Section 3.3 proposes this as the benefit; support is only the ablation gain, not an independent mechanism.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MPSI: Mamba enhancement model for pixel-wise sequential interaction Image Super-Resolution." pith.science (2026). https://pith.science/paper/MNJCQIB5

@misc{pith2026241207222,
  author       = {Pith},
  title        = {Pith review of: MPSI: Mamba enhancement model for pixel-wise sequential interaction Image Super-Resolution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MNJCQIB5}},
  note         = {Machine review of arXiv:2412.07222}
}
read the original abstract

Single image super-resolution (SR) has long posed a challenge in the field of computer vision. While the advent of deep learning has led to the emergence of numerous methods aimed at tackling this persistent issue, the current methodologies still encounter challenges in modeling long sequence information, leading to limitations in effectively capturing the global pixel interactions. To tackle this challenge and achieve superior SR outcomes, we propose the Mamba pixel-wise sequential interaction network (MPSI), aimed at enhancing the establishment of long-range connections of information, particularly focusing on pixel-wise sequential interaction. We propose the Channel-Mamba Block (CMB) to capture comprehensive pixel interaction information by effectively modeling long sequence information. Moreover, in the existing SR methodologies, there persists the issue of the neglect of features extracted by preceding layers, leading to the loss of valuable feature information. While certain existing models strive to preserve these features, they frequently encounter difficulty in establishing connections across all layers. To overcome this limitation, MPSI introduces the Mamba channel recursion module (MCRM), which maximizes the retention of valuable feature information from early layers, thereby facilitating the acquisition of pixel sequence interaction information from multiple-level layers. Through extensive experimentation, we demonstrate that MPSI outperforms existing super-resolution methods in terms of image reconstruction results, attaining state-of-the-art performance.

Figures

Figures reproduced from arXiv: 2412.07222 by the authors.

Figure 1
Figure 1. Architecture of MPSI net work and SAMG. single-image super-resolution tasks, yielding commendable results. DaViT [8] in￾troduced a Transformer architecture amalgamating spatial window self-attention and channel group self-attention to facilitate comprehensive information capture for addressing visual tasks. Building upon DaViT’s framework, Dual Aggrega￾tion Transformer (DAT) [6] devised Dual Spatial Transformer and … view at source ↗
Figure 2
Figure 2. Components of SAMG. The STB and CMB are the components of SAMB. Figures b, c, and d show the structures of STB, CMB, and MCRM respectively. feature sequence does not require position embedding. Mamba has shown great effectiveness and has very good performance in natural language processing. In addition to natural language processing, Mamba has begun to find applica￾tions in various fields, including graph processing… view at source ↗
Figure 3
Figure 3. Architecture of DDBM. The initial value of Ht−1 is a zero matrix. N is the number of features in the feature sequence. batch size, H represents the image height, and W signifies the image width. The convolutional layer within the shallow feature extraction generates features FS ∈ R B×C×H×W , with C denoting the number of channels. In the subsequent stage, the shallow feature is input into the deep feature ex￾tractio… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Visual comparison for single image SR (×4 upscaling). We chose several results to conduct a comparative analysis of the restoration of details by each model. The results of SwinIR, ELAN, and DAT are the lightweight version of the model. performance in the Urban100 data…
Figure 5
Figure 5. Figure 5: Difference maps of ablation studies on the effect of CMB and MCRM. pooling to adjust the feature channels (MRP → None). The results are shown in [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Difference map of ablation studies on DDBM and Mamba recursive process. nel self-attention, both PSNR and SIMM dropped significantly. The comparison between the third column and the fifth column in [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 40 canonical work pages

  1. [1]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops

    Agustsson, E., Timofte, R.: Ntire 2017 challenge on single image super-resolution: Dataset and study. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops. pp. 126–135 (2017) 9, 12

    work page 2017

  2. [2]

    In: Proceedings of the European conference on computer vision (ECCV)

    Ahn, N., Kang, B., Sohn, K.A.: Fast, accurate, and lightweight super-resolution with cascading residual network. In: Proceedings of the European conference on computer vision (ECCV). pp. 252–268 (2018) 10

    work page 2018

  3. [3]

    arXiv preprint arXiv:2007.01127 (2020) 3

    Alaparthi, S., Mishra, M.: Bidirectional encoder representations from transformers (bert): A sentiment analysis odyssey. arXiv preprint arXiv:2007.01127 (2020) 3

    arXiv 2020

  4. [4]

    arXiv preprint arXiv:2402.08678 (2024) 5

    Behrouz, A., Hashemi, F.: Graph mamba: Towards learning on graphs with state space models. arXiv preprint arXiv:2402.08678 (2024) 5

    arXiv 2024

  5. [5]

    Bevilacqua, M., Roumy, A., Guillemot, C., Alberi-Morel, M.L.: Low-complexity single-image super-resolution based on nonnegative neighbor embedding (2012) 9, 10

    work page 2012

  6. [6]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Chen, Z., Zhang, Y., Gu, J., Kong, L., Yang, X., Yu, F.: Dual aggregation trans- former for image super-resolution. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 12312–12321 (2023) 2, 4, 7, 8, 9, 10, 11

    work page 2023

  7. [7]

    Advances in Neural Information Processing Systems 35, 25478–25490 (2022) 1

    Chen, Z., Zhang, Y., Gu, J., Kong, L., Yuan, X., et al.: Cross aggregation trans- former for image restoration. Advances in Neural Information Processing Systems 35, 25478–25490 (2022) 1

    work page 2022

  8. [8]

    In: European Conference on Computer Vision

    Ding, M., Xiao, B., Codella, N., Luo, P., Wang, J., Yuan, L.: Davit: Dual atten- tion vision transformers. In: European Conference on Computer Vision. pp. 74–92. Springer (2022) 4, 7

    work page 2022

Show all 49 references
  1. [9]

    IEEE transactions on pattern analysis and machine intelligence 38(2), 295–307 (2015) 1, 3

    Dong, C., Loy, C.C., He, K., Tang, X.: Image super-resolution using deep convo- lutional networks. IEEE transactions on pattern analysis and machine intelligence 38(2), 295–307 (2015) 1, 3

    2015

  2. [10]

    arXiv preprint arXiv:2010.11929 (2020) 3

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al.: An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020) 3

    2020 arXiv

  3. [11]

    In: 2009 IEEE 12th international conference on computer vision

    Glasner, D., Bagon, S., Irani, M.: Super-resolution from a single image. In: 2009 IEEE 12th international conference on computer vision. pp. 349–356. IEEE (2009) 1

    2009

  4. [12]

    arXiv preprint arXiv:2312.00752 (2023) 2, 4, 8

    Gu, A., Dao, T.: Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752 (2023) 2, 4, 8

    2023 arXiv

  5. [13]

    arXiv preprint arXiv:2111.00396 (2021) 4

    Gu, A., Goel, K., Ré, C.: Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396 (2021) 4

    2021 arXiv

  6. [14]

    Advances in neural information processing systems34, 572–585 (2021) 4

    Gu, A., Johnson, I., Goel, K., Saab, K., Dao, T., Rudra, A., Ré, C.: Combining recurrent, convolutional, and continuous-time models with linear state space layers. Advances in neural information processing systems34, 572–585 (2021) 4

    2021

  7. [15]

    He,K.,Zhang,X.,Ren,S.,Sun,J.:Deepresiduallearningforimagerecognition.In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 770–778 (2016) 2, 3

    2016

  8. [16]

    arXiv preprint arXiv:2402.12192 (2024) 5

    He,X.,Cao,K.,Yan,K.,Li,R.,Xie,C.,Zhang,J.,Zhou,M.:Pan-mamba:Effective pan-sharpening with state space model. arXiv preprint arXiv:2402.12192 (2024) 5

    2024 arXiv

  9. [17]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Hu, J., Shen, L., Sun, G.: Squeeze-and-excitation networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 7132–7141 (2018) 2 Abbreviated paper title 17

    2018

  10. [18]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Huang, G., Liu, Z., Van Der Maaten, L., Weinberger, K.Q.: Densely connected convolutional networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 4700–4708 (2017) 2

    2017

  11. [19]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Huang, J.B., Singh, A., Ahuja, N.: Single image super-resolution from transformed self-exemplars. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 5197–5206 (2015) 9, 10, 12

    2015

  12. [20]

    In: Proceedings of the 27th acm international conference on multimedia

    Hui, Z., Gao, X., Yang, Y., Wang, X.: Lightweight image super-resolution with in- formation multi-distillation network. In: Proceedings of the 27th acm international conference on multimedia. pp. 2024–2032 (2019) 10, 11

    2019

  13. [21]

    arXiv preprint arXiv:1909.05858 (2019) 3

    Keskar, N.S., McCann, B., Varshney, L.R., Xiong, C., Socher, R.: Ctrl: A con- ditional transformer language model for controllable generation. arXiv preprint arXiv:1909.05858 (2019) 3

    2019 arXiv

  14. [22]

    nature 521(7553), 436–444 (2015) 1

    LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. nature 521(7553), 436–444 (2015) 1

    2015

  15. [23]

    In: European Conference on Computer Vision

    Li, W., Cao, Z., Feng, J., Zhou, J., Lu, J.: Label2label: A language modeling framework for multi-attribute learning. In: European Conference on Computer Vision. pp. 562–579. Springer (2022) 3

    2022

  16. [24]

    Advances in Neural Information Processing Systems33, 20343–20355 (2020) 10

    Li, W., Zhou, K., Qi, L., Jiang, N., Lu, J., Jia, J.: Lapar: Linearly-assembled pixel-adaptive regression network for single image super-resolution and beyond. Advances in Neural Information Processing Systems33, 20343–20355 (2020) 10

    2020

  17. [25]

    In: European Confer- ence on Computer Vision

    Li, X., Liu, J., Chen, Z., Zou, Y., Ma, L., Fan, X., Liu, R.: Contourlet residual for prompt learning enhanced infrared image super-resolution. In: European Confer- ence on Computer Vision. pp. 270–288. Springer (2025) 3

    2025

  18. [26]

    arXiv preprint arXiv:2401.00421 (2023) 3

    Li, X., Zou, Y., Liu, J., Jiang, Z., Ma, L., Fan, X., Liu, R.: From text to pixels: A context-aware semantic synergy solution for infrared and visible image fusion. arXiv preprint arXiv:2401.00421 (2023) 3

    2023 arXiv

  19. [27]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Liang, J., Cao, J., Sun, G., Zhang, K., Van Gool, L., Timofte, R.: Swinir: Image restoration using swin transformer. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 1833–1844 (2021) 1, 3, 10, 11

    2021

  20. [28]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops

    Lim,B.,Son,S.,Kim,H.,Nah,S.,MuLee,K.:Enhanceddeepresidualnetworksfor single image super-resolution. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops. pp. 136–144 (2017) 9, 12

    2017

  21. [29]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops

    Lim,B.,Son,S.,Kim,H.,Nah,S.,MuLee,K.:Enhanceddeepresidualnetworksfor single image super-resolution. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops. pp. 136–144 (2017) 10, 11

    2017

  22. [30]

    Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., Guo, B.: Swin transformer:Hierarchical visiontransformerusingshiftedwindows.In:Proceedings of the IEEE/CVF international conference on computer vision. pp. 10012–10022 (2021) 3

    2021

  23. [31]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Lu, Z., Li, J., Liu, H., Huang, C., Zhang, L., Zeng, T.: Transformer for single image super-resolution. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 457–466 (2022) 1, 10, 11

    2022

  24. [32]

    In: Proceedings Eighth IEEE International Conference on Computer Vision

    Martin, D., Fowlkes, C., Tal, D., Malik, J.: A database of human segmented natu- ral images and its application to evaluating segmentation algorithms and measur- ing ecological statistics. In: Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001. vol....

    2001

  25. [33]

    Multimedia Tools and Applications 76, 21811–21838 (2017) 9, 10, 12 18 Y

    Matsui, Y., Ito, K., Aramaki, Y., Fujimoto, A., Ogawa, T., Yamasaki, T., Aizawa, K.: Sketch-based manga retrieval using manga109 dataset. Multimedia Tools and Applications 76, 21811–21838 (2017) 9, 10, 12 18 Y. He et al

    2017

  26. [34]

    Advances in neural information processing sys- tems 32 (2019) 9

    Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., et al.: Pytorch: An imperative style, high- performance deep learning library. Advances in neural information processing sys- tems 32 (2019) 9

    2019

  27. [35]

    Radford, A., Narasimhan, K., Salimans, T., Sutskever, I., et al.: Improving lan- guage understanding by generative pre-training (2018) 3

    2018

  28. [36]

    In: Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, Oc- tober 5-9, 2015, Proceedings, Part III 18

    Ronneberger, O., Fischer, P., Brox, T.: U-net: Convolutional networks for biomed- ical image segmentation. In: Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, Oc- tober 5-9, 2015, Proceedings, Part III 18....

    2015

  29. [37]

    In:ProceedingsoftheIEEEconferenceoncomputervisionandpatternrecognition

    Sajjadi, M.S., Vemulapalli, R., Brown, M.: Frame-recurrent video super-resolution. In:ProceedingsoftheIEEEconferenceoncomputervisionandpatternrecognition. pp. 6626–6634 (2018) 1

    2018

  30. [38]

    In: ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

    Subakan, C., Ravanelli, M., Cornell, S., Bronzi, M., Zhong, J.: Attention is all you need in speech separation. In: ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). pp. 21–25. IEEE (2021) 2, 3

    2021

  31. [39]

    IEEE Transactions on Multimedia23, 1489–1502 (2020) 3

    Tian, C., Xu, Y., Zuo, W., Zhang, B., Fei, L., Lin, C.W.: Coarse-to-fine cnn for image super-resolution. IEEE Transactions on Multimedia23, 1489–1502 (2020) 3

    2020

  32. [40]

    Knowledge-Based Systems205, 106235 (2020) 3

    Tian, C., Zhuge, R., Wu, Z., Xu, Y., Zuo, W., Chen, C., Lin, C.W.: Lightweight image super-resolution with enhanced cnn. Knowledge-Based Systems205, 106235 (2020) 3

    2020

  33. [41]

    In: 2023 IEEE Interna- tional Conference on Image Processing (ICIP)

    Wang, Z., Su, M., Zheng, J.Q., Liu, Y.: Densely connected swin-unet for multiscale information aggregation in medical image segmentation. In: 2023 IEEE Interna- tional Conference on Image Processing (ICIP). pp. 940–944. IEEE (2023) 5

    2023

  34. [42]

    In: Curves and Surfaces: 7th International Conference, Avignon, France, June 24-30, 2010, Revised Selected Papers 7

    Zeyde, R., Elad, M., Protter, M.: On single image scale-up using sparse- representations. In: Curves and Surfaces: 7th International Conference, Avignon, France, June 24-30, 2010, Revised Selected Papers 7. pp. 711–730. Springer (2012) 9, 10

    2012

  35. [43]

    IEEE Transactions on Pattern Analysis and Machine Intelligence44(10), 6360–6376 (2021) 1

    Zhang, K., Li, Y., Zuo, W., Zhang, L., Van Gool, L., Timofte, R.: Plug-and-play image restoration with deep denoiser prior. IEEE Transactions on Pattern Analysis and Machine Intelligence44(10), 6360–6376 (2021) 1

    2021

  36. [44]

    In: European Conference on Computer Vision

    Zhang, X., Zeng, H., Guo, S., Zhang, L.: Efficient long-range attention network for image super-resolution. In: European Conference on Computer Vision. pp. 649–

  37. [45]

    In: Proceedings of the European conference on computer vision (ECCV)

    Zhang, Y., Li, K., Li, K., Wang, L., Zhong, B., Fu, Y.: Image super-resolution using very deep residual channel attention networks. In: Proceedings of the European conference on computer vision (ECCV). pp. 286–301 (2018) 1

    2018

  38. [46]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Zhang, Y., Tian, Y., Kong, Y., Zhong, B., Fu, Y.: Residual dense network for image super-resolution. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2472–2481 (2018) 10

    2018

  39. [47]

    In: Computer Vision–ECCV 2020 Workshops: Glasgow, UK, August 23–28, 2020, Proceedings, Part III 16

    Zhao, H., Kong, X., He, J., Qiao, Y., Dong, C.: Efficient image super-resolution using pixel attention. In: Computer Vision–ECCV 2020 Workshops: Glasgow, UK, August 23–28, 2020, Proceedings, Part III 16. pp. 56–72. Springer (2020) 3

    2020

  40. [48]

    arXiv preprint arXiv:2411.12530 (2024) 3

    Zou, Y., Chen, Z., Zhang, Z., Li, X., Ma, L., Liu, J., Wang, P., Zhang, Y.: Con- tourlet refinement gate framework for thermal spectrum distribution regularized infrared image super-resolution. arXiv preprint arXiv:2411.12530 (2024) 3

    2024 arXiv

  41. [667]

    Springer (2022) 1, 2, 10

    2022

Pith tools

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