pith. sign in

arxiv: 2605.17588 · v1 · pith:ZRJP3TJVnew · submitted 2026-05-17 · 💻 cs.CV

MSIQ: Moment-based Scale-Invariant Quality Measure for Single Image Super-Resolution

Leonid Bedratyuk This is my paper

Pith reviewed 2026-05-20 14:12 UTC · model grok-4.3

classification 💻 cs.CV
keywords scale-invariant quality measuregeometric momentssingle image super-resolutionimage quality assessmentgeometric fidelitySISR evaluationmoment-based metric
0
0 comments X

The pith

MSIQ compares super-resolution images of different sizes using normalized central geometric moments to assess geometric fidelity without resizing.

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

This paper introduces MSIQ as a diagnostic quality measure for single image super-resolution based on comparing normalized central geometric moments. It allows direct evaluation of how well geometric structure is preserved even when the reference and output images have different spatial resolutions. The approach avoids the interpolation errors that come from forcing both images to the same size. Experiments show the measure stays stable under uniform scaling and responds more to geometric changes than to non-geometric ones like JPEG compression. The result holds across a range of super-resolution methods, including various neural network architectures.

Core claim

The paper establishes that comparing normalized central geometric moments produces a deterministic, model-free, analytical, scale-invariant quality score for single image super-resolution. This score directly evaluates preservation of geometric structure and exhibits geometric specificity by separating deformations from non-geometric artifacts. Its response to structural perturbations remains consistent regardless of the super-resolution algorithm used.

What carries the argument

Comparison of normalized central geometric moments between the reference image and the super-resolved image to compute a scale-invariant quality score.

If this is right

  • Quality assessment becomes possible without introducing external interpolation errors from forced resizing.
  • Geometric deformations can be evaluated separately from artifacts such as compression in SR results.
  • The measure gives stable readings across both classical and deep-learning super-resolution methods.
  • It supplies a complementary diagnostic for applications that prioritize geometric accuracy over perceptual similarity.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Evaluation benchmarks for super-resolution could incorporate direct multi-resolution checks as a standard step.
  • The separation of tracking ability from geometric specificity may apply to quality measures in other scale-sensitive imaging tasks.
  • Higher-order moments could be tested as an extension to increase sensitivity to finer geometric details.

Load-bearing premise

Normalized central geometric moments capture the preservation of geometric structure sufficiently and selectively, independent of the super-resolution algorithm.

What would settle it

Uniformly scale an image and recompute MSIQ; a large change in score would contradict scale-invariance, while a geometric warp that should alter structure but leaves the score unchanged would contradict geometric selectivity.

read the original abstract

Assessing the quality of single image super-resolution (SISR) results remains an open methodological problem. Common full-reference metrics (PSNR, SSIM, LPIPS) do not explicitly evaluate the preservation of the geometric structure of images, which is critical for the correctness of scale-based reconstruction. In addition, they require the forced alignment of images to the same size (\textit{forced resizing}), which introduces an external interpolation error into the evaluation process. This paper proposes a diagnostic scale-invariant quality measure, MSIQ (\textit{Moment-based Scale-Invariant Quality}), based on the comparison of normalized central geometric moments of two images. MSIQ enables direct comparison of images with different spatial resolutions without resizing, is mathematically deterministic (\textit{model-free}), and has an analytical form. To provide a theoretical basis for the approach, we introduce a conceptual distinction between the ability of metrics to monotonically track degradation (\textit{tracking ability}) and their geometric selectivity (\textit{geometric specificity}). The experimental validation confirmed the stability of MSIQ under uniform scaling and, at the same time, revealed the high sensitivity of traditional metrics to the choice of interpolation method. The results show that MSIQ has pronounced geometric selectivity: the proposed measure effectively separates geometric deformations from non-geometric artifacts, in particular JPEG compression, unlike pixel-based and perceptual metrics. It is also shown that the response of MSIQ to structural perturbations remains stable across different classes of SR algorithms, including DNN models with different architectures. The proposed measure is a complementary diagnostic tool for domains where geometric fidelity has priority, in particular medical imaging and remote sensing.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript proposes MSIQ, a diagnostic scale-invariant quality measure for single-image super-resolution based on direct comparison of normalized central geometric moments between a reference and a super-resolved image. It claims to be mathematically deterministic and model-free with an analytical form, to enable comparison across different spatial resolutions without forced resizing or interpolation, and to exhibit geometric selectivity by separating geometric deformations from non-geometric artifacts such as JPEG compression while remaining stable across SR algorithm classes.

Significance. If the central claims hold, MSIQ would serve as a useful complementary diagnostic for applications that prioritize geometric fidelity (medical imaging, remote sensing). The explicit distinction between monotonic tracking ability and geometric specificity is a constructive conceptual contribution, and the avoidance of resizing-induced interpolation error addresses a practical limitation of existing full-reference metrics.

major comments (1)
  1. [Experimental validation] The central claim of geometric selectivity rests on the assertion that normalized central moments (defined via the standard integral form η_pq = μ_pq / μ_00^{(p+q)/2+1}) remain sensitive to structural perturbations while being insensitive to non-geometric artifacts. Because these moments are global integrals over the entire support, a localized affine warp or edge misalignment (typical SISR failure mode) occupying only a small area fraction can be offset elsewhere, leaving low-order moments nearly unchanged. The experimental validation section should therefore include quantitative tests with spatially confined perturbations whose area fraction is varied, together with the resulting moment deviation curves, to substantiate that the chosen moment orders and comparison operator retain sensitivity under realistic localized distortions.
minor comments (2)
  1. Specify the exact moment orders (p,q) retained in the MSIQ vector and the precise distance or similarity operator used to compare the two moment sets; without this the analytical form remains incomplete.
  2. Clarify whether the reported stability under uniform scaling was verified analytically (via the normalization) or only empirically, and state the range of scale factors tested.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comments and the positive evaluation of MSIQ's potential utility in applications prioritizing geometric fidelity. We address the single major comment below and will incorporate the requested validation in the revised manuscript.

read point-by-point responses

  1. Referee: [Experimental validation] The central claim of geometric selectivity rests on the assertion that normalized central moments (defined via the standard integral form η_pq = μ_pq / μ_00^{(p+q)/2+1}) remain sensitive to structural perturbations while being insensitive to non-geometric artifacts. Because these moments are global integrals over the entire support, a localized affine warp or edge misalignment (typical SISR failure mode) occupying only a small area fraction can be offset elsewhere, leaving low-order moments nearly unchanged. The experimental validation section should therefore include quantitative tests with spatially confined perturbations whose area fraction is varied, together with the resulting moment deviation curves, to substantiate that the chosen moment orders and comparison operator retain sensitivity under realistic localized distortions.

    Authors: We agree that the global nature of central moments raises a valid question about sensitivity to highly localized perturbations, and that explicit tests with varying area fractions would strengthen the evidence for geometric selectivity. Our current experiments already demonstrate that MSIQ separates geometric deformations from non-geometric artifacts (e.g., JPEG compression) and remains stable across SR algorithm classes, but they do not directly quantify the effect of spatially confined distortions. In the revised manuscript we will add a new set of controlled experiments applying localized affine warps and edge misalignments over controlled area fractions (5 %, 10 %, 15 %, 25 % of the image support). We will report the resulting deviations in the normalized central moments (orders 0–4) and the corresponding MSIQ values, including plots of moment deviation versus perturbation area fraction. These results will be used to confirm that the chosen moment orders and comparison operator retain adequate sensitivity for the scale of distortions typical in SISR. revision: yes

Circularity Check

0 steps flagged

MSIQ defined directly from standard normalized central moments with no circular reduction

full rationale

The paper defines MSIQ explicitly as a comparison of normalized central geometric moments (standard mathematical objects) between images at different resolutions, without any fitting to data subsets or renaming of predictions. The introduced distinction between tracking ability and geometric specificity is a conceptual framing presented within the paper. Experimental validation of scaling stability and selectivity is reported separately as empirical confirmation rather than part of the derivation. No load-bearing self-citations or ansatzes imported from prior author work are indicated in the provided text, so the central construction remains self-contained against external moment theory.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that geometric moments can isolate structural fidelity; no free parameters or invented entities are indicated in the abstract.

axioms (1)
  • domain assumption Normalized central geometric moments capture preservation of geometric structure across different image resolutions
    Invoked to justify direct comparison without resizing and to claim geometric selectivity.

pith-pipeline@v0.9.0 · 5818 in / 1152 out tokens · 48161 ms · 2026-05-20T14:12:02.955441+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

50 extracted references · 50 canonical work pages

  1. [1]

    L. Yue, H. Shen, J. Li, Q. Yuan, H. Zhang, L. Zhang, Image super-resolution: The techniques, applications, and future, Signal Process. 128 (2016) 389–408. https://doi.org/10.1016/j.sigpro.2016.05.002

    work page doi:10.1016/j.sigpro.2016.05.002 2016

  2. [2]

    H. Chen, X. He, L. Qing, Y. Wu, C. Ren, R.E. Sheriff, C. Zhu, Real-world single image super-resolution: A brief review, Inf. Fusion. 79 (2022) 124–145. https://doi.org/10.1016/j.inffus.2021.09.005 1https://github.com/LeonidBed/msiq-metrics MOMENT-BASED SCALE-INV ARIANT QUALITY MEASURE 21

    work page doi:10.1016/j.inffus.2021.09.005 2022

  3. [3]

    Schmon, and Chris G

    A. Lugmayr, M. Danelljan, R. Timofte et al., NTIRE 2022 challenge on learning the super- resolution space, IEEE/CVF Conference on Computer Vision and Pattern Recognition (2022) 786-797. https://doi.org/10.1109/CVPRW56347.2022.00094

    work page doi:10.1109/cvprw56347.2022.00094 2022

  4. [4]

    N. Ahn, B. Kang, K.A. Sohn, Efficient deep neural network for photo-realistic image super-resolution, Pattern Recognit. 127 (2022) 108649. https://doi.org/10.1016/j.patcog.2022.108649

    work page doi:10.1016/j.patcog.2022.108649 2022

  5. [5]

    S. Gao, X. Liu, B. Zeng et al., Implicit diffusion models for continuous super-resolution, IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (2023) 10021–10030. https://doi.org/10.1109/CVPR52729.2023.00966

    work page doi:10.1109/cvpr52729.2023.00966 2023

  6. [6]

    Reibman, R.M

    A.R. Reibman, R.M. Bell, S. Gray, Quality assessment for super-resolution image enhancement, IEEE Int. Conf. Image Process. (2006) 2017–2020. https://doi.org/10.1109/ICIP.2006.312895

    work page doi:10.1109/icip.2006.312895 2006

  7. [7]

    Ma, C.-Y

    C. Ma, C.-Y. Yang, X. Yang, M.-H. Yang, Learning a no-reference quality metric for single-image super- resolution, Comput. Vis. Image Underst. 158 (2017) 1–16. https://doi.org/10.1016/j.cviu.2016.12.009

    work page doi:10.1016/j.cviu.2016.12.009 2017

  8. [8]

    Engeldrum, Psychometric scaling: avoiding the pitfalls and hazards, IS&T’s PICS Conf

    P.G. Engeldrum, Psychometric scaling: avoiding the pitfalls and hazards, IS&T’s PICS Conf. Proc. (2001) 101–107

    work page 2001

  9. [9]

    de Ridder, S

    H. de Ridder, S. Endrikhovski, Image quality is fun: reflections on fidelity, usefulness and naturalness, SID Symp. Dig. Tech. Pap. 33 (2002) 986–989. https://doi.org/10.1889/1.1830949

    work page doi:10.1889/1.1830949 2002

  10. [10]

    Bianco, G

    S. Bianco, G. Ciocca, F. Marini, R. Schettini, Image quality assessment by preprocessing and full reference model combination, Proc. SPIE. 7242 (2009) 232-240. https://doi.org/10.1117/12.806693

    work page doi:10.1117/12.806693 2009

  11. [11]

    Bovik, Hamid R

    Z. Wang, A.C. Bovik, H.R. Sheikh, E.P. Simoncelli, Image quality assessment: from error visibility to struc- tural similarity, IEEE Trans. Image Process. 13 (2004) 600–612. https://doi.org/10.1109/TIP.2003.819861

    work page doi:10.1109/tip.2003.819861 2004

  12. [12]

    Sheikh, A

    H. Sheikh, A. Bovik, Image information and visual quality, IEEE Trans. Image Process. 15(2) (2006) 430–444. https://doi.org/10.1109/TIP.2005.859378

    work page doi:10.1109/tip.2005.859378 2006

  13. [13]

    Efros, Eli Shechtman, and Oliver Wang

    R. Zhang, P. Isola, A.A. Efros, E. Shechtman, O. Wang, The unreasonable effectiveness of deep features as a perceptual metric, IEEE Conf. Comput. Vis. Pattern Recognit. (2018) 586–595. https://doi.org/10.1109/CVPR.2018.00068

    work page doi:10.1109/cvpr.2018.00068 2018

  14. [14]

    K. Ding, K. Ma, S. Wang and E. P. Simoncelli, Image Quality Assessment: Unifying Structure and Texture Similarity, in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 5, pp. 2567- 2581,(2022). https://doi.org/10.1109/TPAMI.2020.3045810

    work page doi:10.1109/tpami.2020.3045810 2022

  15. [15]

    Hu, Visual pattern recognition by moment invariants, IRE Trans

    M.K. Hu, Visual pattern recognition by moment invariants, IRE Trans. Inf. Theory. 8 (1962) 179–187. https://doi.org/10.1109/TIT.1962.1057692

    work page doi:10.1109/tit.1962.1057692 1962

  16. [16]

    Flusser, T

    J. Flusser, T. Suk, B. Zitova, 2D and 3D Image Analysis by Moments, John Wiley & Sons, Chichester (2017). https://doi.org/10.1002/9781119039402

    work page doi:10.1002/9781119039402 2017

  17. [17]

    W. Zhou, Z. Wang, Quality assessment of image super-resolution: balancing deterministic and statistical fidelity, Proc. 30th ACM Int. Conf. Multimedia (2022) 934–942. https://doi.org/10.1145/3503161.3547899

    work page doi:10.1145/3503161.3547899 2022

  18. [18]

    Lyapustin, A

    E. Lyapustin, A. Kirillova, V. Meshchaninov et al., Towards true detail restoration for super-resolution: A benchmark and a quality metric, arXiv prepr. arXiv:2203.08923 (2022). https://arxiv.org/abs/2203.08923

    work page arXiv 2022

  19. [19]

    Wang, E.Y

    M. Wang, E.Y. Wang, G.H. Pan, Image quality assessment based on invariant moments similarity, Adv. Mater. Res. 546 (2012) 565–569. https://doi.org/10.4028/www.scientific.net/AMR.546-547.565

    work page doi:10.4028/www.scientific.net/amr.546-547.565 2012

  20. [20]

    Premaratne, M

    P. Premaratne, M. Premaratne, Image similarity index based on moment invariants of approximation level of discrete wavelet transform, Electron. Lett. 48 (2012) 1465–1467. https://doi.org/10.1049/el.2012.2739

    work page doi:10.1049/el.2012.2739 2012

  21. [21]

    Saharia, J

    C. Saharia, J. Ho, W. Chan, T. Salimans, D.J. Fleet, M. Norouzi, Image super-resolution via iterative refinement, IEEE Trans. Pattern Anal. Mach. Intell. 45 (2023) 4713–4726. https://doi.org/10.1109/TPAMI.2022.3204461

    work page doi:10.1109/tpami.2022.3204461 2023

  22. [22]

    R. Wu, T. Yang, L. Sun et al., SeeSR: Towards semantics-aware real-world image super-resolution, IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (2024) 25456–25467. https://doi.org/10.48550/arXiv.2311.16518

    work page doi:10.48550/arxiv.2311.16518 2024

  23. [23]

    J. Wang, Z. Yue, S. Zhou, K.C. Chan, C.C. Loy, Exploiting diffusion prior for real-world image super- resolution, Int. J. Comput. Vis. 132 (2024) 5929–5949. https://doi.org/10.1007/s11263-024-02168-7

    work page doi:10.1007/s11263-024-02168-7 2024

  24. [24]

    Arabboev, S

    M. Arabboev, S. Begmatov, M. Rikhsivoev, K. Nosirov, S. Saydiakbarov, A comprehensive review of image super-resolution metrics: Classical and AI-based approaches, Acta IMEKO. 13(1) (2024) 1–8. https://doi.org/10.21014/actaimeko.v13i1.1679

    work page doi:10.21014/actaimeko.v13i1.1679 2024

  25. [25]

    Batini, M

    C. Batini, M. Scannapieco, Data and information quality, Springer International Publishing, Cham (2016) 21-51. https://doi.org/10.1007/978-3-319-24106-7

    work page doi:10.1007/978-3-319-24106-7 2016

  26. [26]

    Y. Blau, T. Michaeli, The perception-distortion tradeoff, IEEE Conf. Comput. Vis. Pattern Recognit. (2018) 6228–6237. https://doi.org/10.1109/CVPR.2018.00652 22 LEONID BEDRATYUK

    work page doi:10.1109/cvpr.2018.00652 2018

  27. [27]

    Jiang, Z

    Q. Jiang, Z. Liu, K. Gu, F. Shao, X. Zhang, H. Liu, W. Lin, Single image super-resolution quality assess- ment: A real-world dataset, subjective studies, and an objective metric, IEEE Trans. Image Process. 31 (2022) 2279–2294. https://doi.org/10.1109/TIP.2022.3154588

    work page doi:10.1109/tip.2022.3154588 2022

  28. [28]

    Greeshma, V.R

    M.S. Greeshma, V.R. Bindu, Super-resolution Quality Criterion (SRQC): A super-resolution image quality assessment metric, Multimed. Tools Appl. 79 (2020) 35125–35146. https://doi.org/10.1007/s11042-020- 09352-0

    work page doi:10.1007/s11042-020- 2020

  29. [29]

    D. Fan, K. Zhang, X. Quan, Y. Meng, No-reference super-resolution image quality assessment based on upscaling-factor aware contrastive learning, J. Northwest Univ. (Nat. Sci. Ed.). 55 (2025) 309–319. https://doi.org/10.16152/j.cnki.xdxbzr.2025-02-008

    work page doi:10.16152/j.cnki.xdxbzr.2025-02-008 2025

  30. [30]

    L. Shu, Q. Zhu, Y. He, W. Chen, J. Yan, A survey of super-resolution image quality assessment, Neuro- computing. 621 (2025) 129279. https://doi.org/10.1016/j.neucom.2024.129279

    work page doi:10.1016/j.neucom.2024.129279 2025

  31. [31]

    S. Su, J.M. Rocafort, D. Xue et al., Rethinking image evaluation in super-resolution, arXiv prepr. arXiv:2503.13074 (2025). https://doi.org/10.48550/arXiv.2503.13074

    work page doi:10.48550/arxiv.2503.13074 2025

  32. [32]

    Majlessi, A.M

    K. Majlessi, A.M. Soltani, M.E. Mahdavi et al., Self-supervised image super-resolution quality assessment based on content-free multi-model oriented representation learning, arXiv prepr. arXiv:2602.10744 (2026). https://doi.org/10.48550/arXiv.2602.10744

    work page doi:10.48550/arxiv.2602.10744 2026

  33. [33]

    arXiv:2504.18524 (2025)

    F.Zhang, S.B.Rangrej, T.Aumentado-Armstrongetal., Augmentingperceptualsuper-resolutionviaimage quality predictors, arXiv prepr. arXiv:2504.18524 (2025). https://doi.org/10.48550/arXiv.2504.18524

    work page doi:10.48550/arxiv.2504.18524 2025

  34. [34]

    Zhang, W

    K. Zhang, W. Chen, T. Zhao and Z. Wang, Structural Similarity in Deep Features: Unified Image Quality Assessment Robust to Geometrically Disparate Reference, in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 48, no. 3, pp. 2581-2595,(2026), https://doi.org/10.1109/TPAMI.2025.3627285

    work page doi:10.1109/tpami.2025.3627285 2026

  35. [35]

    Reisenhofer, S

    R. Reisenhofer, S. Bosse, G. Kutyniok, T. Wiegand, A Haar wavelet-based perceptual sim- ilarity index for image quality assessment, Signal Process. Image Commun. 61 (2018) 33–43. https://doi.org/10.1016/j.image.2017.11.001

    work page doi:10.1016/j.image.2017.11.001 2018

  36. [36]

    Nafchi, A

    H.Z. Nafchi, A. Shahkolaei, R. Hedjam, M. Cheriet, Mean deviation similarity index: Ef- ficient and reliable full-reference image quality evaluator, IEEE Access. 4 (2016) 5579–5590. https://doi.org/10.1109/ACCESS.2016.2604042

    work page doi:10.1109/access.2016.2604042 2016

  37. [37]

    W. Xue, L. Zhang, X. Mou, A.C. Bovik, Gradient magnitude similarity deviation: A highly efficient perceptual image quality index, IEEE Trans. Image Process. 23(2) (2014) 684–695. https://doi.org/10.1109/TIP.2013.2293423

    work page doi:10.1109/tip.2013.2293423 2014

  38. [38]

    Flusser, T

    J. Flusser, T. Suk, Pattern recognition by affine moment invariants, Pattern Recognit. 26 (1993) 167–174. https://doi.org/10.1016/0031-3203(93)90098-H

    work page doi:10.1016/0031-3203(93)90098-h 1993

  39. [39]

    Flusser, T

    J. Flusser, T. Suk, A moment-based approach to registration of satellite images, IEEE Trans. Geosci. Remote Sens. 32 (1994) 382–387. https://doi.org/10.1109/36.295052

    work page doi:10.1109/36.295052 1994

  40. [40]

    C.Y. Wee, R. Paramesran, R. Mukundan, Quality assessment of Gaussian blurred im- ages using symmetric geometric moments, Int. Conf. Image Anal. Process. (2007) 805–812. https://doi.org/10.1109/ICIAP.2007.4362875

    work page doi:10.1109/iciap.2007.4362875 2007

  41. [41]

    C.Y. Wee, R. Paramesran, R. Mukundan, X. Jiang, Image quality assessment by discrete orthogonal moments, Pattern Recognit. 43 (2010) 4055–4068. https://doi.org/10.1016/j.patcog.2010.05.026

    work page doi:10.1016/j.patcog.2010.05.026 2010

  42. [42]

    Premaratne, P., & Premaratne, M. (2014). Image matching using moment invariants. Neurocomputing, 137, 65-70. https://doi.org/10.1016/j.neucom.2013.02.058

    work page doi:10.1016/j.neucom.2013.02.058 2014

  43. [43]

    Singh, A.N

    C. Singh, A.N. Aggarwal, Single-image super-resolution using orthogonal rotation invariant moments, Com- put. Electr. Eng. 62 (2017) 266–280. 10.1016/j.compeleceng.2017.02.009

    work page doi:10.1016/j.compeleceng.2017.02.009 2017

  44. [44]

    Kanan, M

    H.R. Kanan, M. Nazari, H.R. Pourreza, Rotation invariant multi-frame image super resolution reconstruc- tion using Pseudo Zernike Moments, Signal Process. 118 (2015) 103–114. 10.1016/j.sigpro.2015.05.015

    work page doi:10.1016/j.sigpro.2015.05.015 2015

  45. [45]

    Bedratyuk, H.I

    L.P. Bedratyuk, H.I. Bedratyuk, Scaling image quality assessment via normalized moment invariants, Meas. Comput. Devices Technol. Process. 1 (2020) 51–59

    work page 2020

  46. [46]

    L. Bedratyuk, 2D moment invariants from the point of view of the classical invariant theory, Journal of Mathematical Imaging and Vision, 62(8), (2020), 1062-1075 https://doi.org/10.1007/s10851-020-00954-9

    work page doi:10.1007/s10851-020-00954-9 2020

  47. [47]

    Dong, C.C

    C. Dong, C.C. Loy, X. Tang, Accelerating the super-resolution convolutional neural network, Eur. Conf. Comput. Vis. (2016) 391–407

    work page 2016

  48. [48]

    W. Shi, J. Caballero, F. Huszar et al., Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network, IEEE Conf. Comput. Vis. Pattern Recognit. (2016) 1874–1883

    work page 2016

  49. [49]

    Lai, J.B

    W.S. Lai, J.B. Huang, N. Ahuja, M.H. Yang, Deep Laplacian pyramid networks for fast and accurate super-resolution, IEEE Conf. Comput. Vis. Pattern Recognit. (2017) 624–632. https://doi.org/10.1109/CVPR.2017.618 MOMENT-BASED SCALE-INV ARIANT QUALITY MEASURE 23

    work page doi:10.1109/cvpr.2017.618 2017

  50. [50]

    B. Lim, S. Son, H. Kim, S. Nah, K.M. Lee, Enhanced deep residual networks for single im- age super-resolution, IEEE Conf. Comput. Vis. Pattern Recognit. Workshops (2017) 136–144. https://doi.org/10.1109/CVPRW.2017.151 Khmelnytsky National University, F aculty of Information Technology, Ukraine Email address:leonidbedratyuk@khmnu.edu.ua

    work page doi:10.1109/cvprw.2017.151 2017