Pith. sign in

REVIEW 3 major objections 5 minor 80 references

A Systematic Investigation on Deep Learning-Based Omnidirectional Image and Video Super-Resolution

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper argues that deep learning-based omnidirectional image and video super-resolution methods, which perform well on synthetic bicubic-downsampling benchmarks, generalize poorly to authentically captured degradations, and it…

desk verdict Useful survey and a benchmark idea worth taking seriously, but the real-world generalization claim is not yet supported by the NR-IQA-only evaluation. read the letter →

arxiv 2506.06710 v1 pith:RYWNUPMH submitted 2025-06-07 cs.CV eess.IV

Qianqian Zhao , Chunle Guo , Tianyi Zhang , Junpei Zhang , Peiyang Jia , Tan Su , Wenjie Jiang , Chongyi Li This is my paper
classification cs.CVeess.IV
keywords omnidirectionalimagesuper-resolutionvideo360Instadatasetreal-worlddegradationno-referencequalityassessmentvirtualrealitydeeplearningpanoramicdistortion
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

This paper is a systematic survey of deep learning-based omnidirectional image and video super-resolution, and its central contribution is a real-world evaluation. The authors argue that standard benchmarks, which create low-resolution panoramas by bicubic or fisheye downsampling, do not represent what happens when 360-degree cameras record actual scenes. To close that gap, they introduce 360Insta, a dataset of 1,500 authentically degraded omnidirectional images plus video, spanning indoor, outdoor, day, night, motion blur, dim lighting, and varied exposure. Current models that look strong on synthetic benchmarks receive uniformly low scores on four no-reference quality metrics when run on 360Insta. The paper concludes that current methods do not generalize well to real-world degradations and that the field needs more realistic benchmarks.

What carries the argument

The load-bearing instrument is 360Insta, a benchmark of 1,500 authentically degraded 360-degree images in equirectangular projection, the common flat panoramic format, at 3840 by 1920 resolution, plus a video set, captured with a consumer 360-degree camera under real conditions and deliberately including multi-scene blur, different lighting, dim conditions, motion blur, and exposure variation. Because 360Insta has no high-resolution ground truth, evaluation runs through four no-reference image quality metrics—NIQE, MUSIQ, MANIQA, and CLIPIQA—computed on 16 blocks of each image after super-resolution at scale factors 2 and 4. The comparison uses the publicly released code of existing ODISR models such as 360-SS, OSRT, BPOSR, and OmniSSR. The argument's force comes from the contrast: the same methods score well under synthetic ERP and fisheye downsampling protocols in Tables 4 and 5 and then produce uniformly low no-reference scores on real captures in Table 8.

What would settle it

A subjective study would settle the claim: show viewers the original low-resolution frames, bicubic upscaling, and each model's scale-factor-2 and scale-factor-4 outputs from a subset of 360Insta on a head-mounted display, then correlate their quality ratings with NIQE, MUSIQ, MANIQA, and CLIPIQA. If human preference ranks the models differently, or rates the super-resolved frames as acceptable quality, the paper's conclusion that current methods do not perform well on real degradation would not stand.

Watch

Extended reading notes

Core claim

The central claim is that deep learning-based omnidirectional super-resolution has been tuned to synthetic degradations and has not been shown to work on real-world panoramic captures. Trained and tested on datasets like ODISR, ODISR-clean, SUN360, and SUN360-clean, where low-resolution inputs are produced by bicubic or fisheye downsampling, the surveyed methods achieve competitive scores on full-reference metrics. On 360Insta, which contains images with motion blur, dim conditions, changing exposure, and varied scenes, the same methods score uniformly low on NIQE, MUSIQ, MANIQA, and CLIPIQA at both scale factors 2 and 4. The paper treats this as evidence that the field requires real-degradation benchmarks and models robust to them, rather than only higher numbers on synthetic pairs.

Load-bearing premise

The real-world generalization conclusion rests on the assumption that the four no-reference quality metrics (NIQE, MUSIQ, MANIQA, and CLIPIQA) faithfully capture perceived quality for distorted 360-degree images; since 360Insta has no high-resolution ground truth, there is no direct check of what the super-resolved outputs actually recover.

Editorial extensions

If this is right

  • On 360Insta, every tested model receives low NIQE, MUSIQ, MANIQA, and CLIPIQA scores at scale factors 2 and 4, so claiming real-world readiness for VR and AR applications from synthetic-benchmark results is not supported by this evaluation.
  • Increasing the upscaling factor makes real-degradation quality worse for the tested models, and no method in the comparison escapes that trend.
  • The survey's unified test protocols and public dataset make it possible for future ODISR and ODVSR work to be measured for real-world robustness rather than only for synthetic fidelity.
  • A method that is strong on synthetic benchmarks is not automatically strong on authentic captures; the rankings change when the evaluation moves to 360Insta.
  • Because 360Insta includes videos, the same real-degradation gap can be checked for video super-resolution, although the paper notes its video set could not be included in the main text.

Reading between the lines

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

  • A natural experiment the paper does not run is to fine-tune or train an ODISR model on real-degradation examples like 360Insta and see whether the no-reference scores improve; if they do, the failure is a distribution-shift problem rather than a hard limit of super-resolution.
  • Because 360Insta has no paired high-resolution ground truth, it cannot distinguish a model that genuinely recovers detail from one that merely changes texture in ways the metrics reward; pairing the benchmark with matched high-quality captures of the same scenes would convert it from a diagnostic set into a trainable one.
  • The authors' conclusion depends on four no-reference metrics, and those metrics were not designed for equirectangular panoramas; a head-mounted-display subjective study comparing super-resolved outputs against bicubic upscaling would test whether viewers actually share the metrics' verdict.
  • The taxonomy in the survey suggests that real-degradation robustness might require combining distortion-aware losses, projection fusion, and temporal alignment, but the paper does not propose such a combination.
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 / 5 minor

Summary. The paper is a systematic survey of deep learning-based omnidirectional image and video super-resolution (ODISR/ODVSR). It taxonomizes existing methods by design principle (distortion maps, projection designs, position encoding, training strategies, diffusion), describes common datasets and metrics, and introduces a new dataset, 360Insta, of authentically degraded omnidirectional images and videos captured with a consumer 360-degree camera. The authors benchmark several published ODISR methods on established synthetic-degradation datasets (ODISR, SUN360 and clean variants) under both ERP and fisheye downsampling, compare model complexity, and evaluate a subset of methods on 360Insta using no-reference IQA metrics. The central empirical claim is that current ODISR methods, which perform well on synthetic bicubic degradation, achieve uniformly low no-reference quality scores on 360Insta, indicating poor generalization to real-world degradations.

Significance. If the 360Insta evaluation is trustworthy, the dataset and the generalization finding would be a useful contribution to the omnidirectional SR community, where most benchmarks rely on synthetic bicubic or fisheye downsampling. The paper also provides a valuable organized testbed: it compiles representative methods, reports results using official implementations, includes parameter/FLOP comparisons, and makes code and data links publicly available (with the caveat noted below). These are concrete strengths. However, the paper's main empirical novelty—the real-world generalization claim—is currently supported only by unvalidated no-reference metrics without a low-resolution baseline, which is a load-bearing weakness. The paper does not derive new theory; the value rests on the reliability of the new benchmark, so the evaluation protocol must be robust.

major comments (3)
  1. [Section 6.2.1, text after Table 4] The central claim that current ODISR methods 'do not perform well on the real degradation dataset' rests entirely on four no-reference metrics (NIQE, MUSIQ, MANIQA, CLIPIQA) applied to 360Insta, which has no paired high-resolution ground truth. This is problematic for two reasons. First, the manuscript itself notes in Section 6.1.2 that 'there is still no universally accepted metric' for this field, and none of the four chosen metrics was developed or validated for equirectangular projection (ERP) content, which has latitude-dependent stretching, seams, and large unpopulated sky/ground regions. Second, Table 8 contains no low-resolution input baseline: it reports only SR outputs, so low NIQE or MANIQA values could reflect the inherently degraded content of 360Insta or systematic metric bias rather than a failure of SR generalization. The evaluation also splits 3840x1920 images into 16 blocks, which can alter natural-scene statistics on which NIQE relies. To make the conclusion load-bearing, the authors should report the same NR-IQA scores for the LR inputs and for a trivial baseline (e.g., bicubic upsampling), and should add at least a small paired real-data subset with full-reference metrics or a human rating study. Without such calibration, Section 6.2.5's conclusion is not supported.
  2. [Section 6.2.5, Table 8] The text states: 'At a scaling factor of 16, the BPOSR method performs well in PSNR and SSIM, reaching 24.01 and 0.6730, respectively.' In Table 4, these exact values (24.01 PSNR, 0.6730 SSIM on ODISR) are the BPOSR x8 results; the actual x16 values are 22.05 PSNR and 0.6258 SSIM. This misattribution affects the discussion of large-scale SR performance and should be corrected, along with the following sentence about WS-PSNR/WS-SSIM, which similarly appears to reference the x8 row.
  3. [Section 5.3 and Section 6.2.6] The dataset contribution is incompletely documented. Section 5.3 gives only image counts for 360Insta and no count, resolution, or duration details for the claimed 360Insta video dataset, and Section 6.2.6 provides the video dataset link as 'https://drive.google.com/drive/my-drive', which is a private placeholder rather than an accessible public resource. Since the abstract and introduction promise that 'all datasets... are publicly available,' this unverifiable availability undermines a core contribution and needs to be fixed before the dataset claim can be accepted.
minor comments (5)
  1. [Section 5.1] The dataset table lists '360HUD' for the video dataset while Section 5.2 and reference [39] use '360UHD'; this inconsistency should be reconciled.
  2. [Section 5.2] There is a typographical error: '2048°×1024' includes a degree symbol where none is intended; it should read '2048×1024'.
  3. [Section 6.2.4] The sentence 'The ODV-SR data set is then divided into training, validation, and testing sets, which include 270 clips, 20 clips, and 25 clips, respectively' is correct, but the adjacent description of ODV360 says 'all videos were downsampled to a resolution of 2K (2160×1080), with each clip consisting of 100 frames'; the numbers for ODV360 (210/20/20) do not match the earlier statement of '90 HR videos' plus '160 videos' collected with Insta360 cameras—please clarify the total counts.
  4. [Section 6.2.6] The table lists a method named 'Aalign' [68]; the standard name in the cited reference is 'A2N' (or 'Align' in some papers). Please use the canonical name to avoid confusion.
  5. [Section 1] In the visual-quality discussion, the text says 'OSRT [23] reveals pronounced artifacts that compromise visual coherence while it still performs well'; this sentence contradicts the stronger quantitative claims about OSRT's superiority in the same section. Please rephrase to state which artifacts are observed and how they interact with the quantitative results.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 360Insta evaluation is empirical benchmarking; the NR-IQA validity concern is an external-validity issue, not a derivation that reduces to its inputs.

full rationale

This paper is a survey plus a new real-world benchmark dataset (360Insta) and an empirical comparison. The strongest claim—that existing ODISR methods trained on synthetic bicubic/fisheye degradations achieve low no-reference quality scores on 360Insta and thus generalize poorly to real-world degradations—is an empirical measurement, not a derivation. The dataset is not defined from the outputs, no parameter is fitted to the 360Insta scores, and no 'prediction' is constructed from the same data used to fit it. The only self-citation is a routine citation to Srformer (ref [15]) in a list of ViT-based SISR methods, which is not load-bearing. The paper itself flags a limitation in Sec. 6.1.2: 'there is still no universally accepted metric in the field.' Sec. 6.2.5 evaluates only NR-IQA metrics without paired ground truth; this is a legitimate threat to the validity of the generalization conclusion, but it is an external-validity/measurement concern, not circularity. Because no equation or argument reduces to its own inputs, the appropriate circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper is a survey and dataset contribution. There are no fitted free parameters or invented theoretical entities. The key assumptions are the representativeness of the captured data, the validity of NR-IQA metrics for the real-world benchmark, and the standard fisheye projection model used in the background.

assumptions (3)
  • domain assumption The Insta360 X4 camera captures omnidirectional content that is representative of real-world 360-degree imagery and its degradations.
    The dataset's validity as a real-world benchmark relies on this representation claim, introduced in Section 5.3.
  • domain assumption No-reference image quality metrics (NIQE, MUSIQ, MANIQA, CLIPIQA) are valid measures of super-resolution quality on distorted 360-degree content.
    Evaluation on 360Insta uses only these metrics because no HR ground truth exists; the conclusion rests on this assumption in Section 6.2.5.
  • standard math The standard fisheye polynomial projection model of Kannala and Brandt holds for the cameras used to capture or synthesize data.
    Section 2 presents the fisheye imaging equations, but the paper does not calibrate this model to the actual Insta360 camera; it is carried over from prior literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Systematic Investigation on Deep Learning-Based Omnidirectional Image and Video Super-Resolution." pith.science (2026). https://pith.science/paper/RYWNUPMH

@misc{pith2026250606710,
  author       = {Pith},
  title        = {Pith review of: A Systematic Investigation on Deep Learning-Based Omnidirectional Image and Video Super-Resolution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RYWNUPMH}},
  note         = {Machine review of arXiv:2506.06710}
}
read the original abstract

Omnidirectional image and video super-resolution is a crucial research topic in low-level vision, playing an essential role in virtual reality and augmented reality applications. Its goal is to reconstruct high-resolution images or video frames from low-resolution inputs, thereby enhancing detail preservation and enabling more accurate scene analysis and interpretation. In recent years, numerous innovative and effective approaches have been proposed, predominantly based on deep learning techniques, involving diverse network architectures, loss functions, projection strategies, and training datasets. This paper presents a systematic review of recent progress in omnidirectional image and video super-resolution, focusing on deep learning-based methods. Given that existing datasets predominantly rely on synthetic degradation and fall short in capturing real-world distortions, we introduce a new dataset, 360Insta, that comprises authentically degraded omnidirectional images and videos collected under diverse conditions, including varying lighting, motion, and exposure settings. This dataset addresses a critical gap in current omnidirectional benchmarks and enables more robust evaluation of the generalization capabilities of omnidirectional super-resolution methods. We conduct comprehensive qualitative and quantitative evaluations of existing methods on both public datasets and our proposed dataset. Furthermore, we provide a systematic overview of the current status of research and discuss promising directions for future exploration. All datasets, methods, and evaluation metrics introduced in this work are publicly available and will be regularly updated. Project page: https://github.com/nqian1/Survey-on-ODISR-and-ODVSR.

Figures

Figures reproduced from arXiv: 2506.06710 by the authors.

Figure 1
Figure 1. Overview of omnidirectional imaging. (a) Omnidirectional imaging and projection. (b) Examples of LR ODIs captured in the real world. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. A review of deep learning-based ODISR (red) and ODVSR (blue) methods. The Distortion Map [23], [38], [39], [40], [41], [42] refers to the fact that the distortion map is used as a condition to guide model training. Projection [27], [28], [43], [44] refers to that the model is designed using multiple projection methods such as hexahedron and cylindrical projection. Training [29], [30] refers to enhancing SR performan… view at source ↗
Figure 3
Figure 3. Fisheye camera imaging principle. Spherical Projection: 3D [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: ERP and CMP projections. (a) Example of spherical projection to ERP. (b) Distortion map of ERP. The increased intensity of color indicates [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Illustrations of the ODIs with different image content in our 360Insta dataset. 360Insta includes different scenes (e.g., nature landscape, [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Visual comparison of the super-resolution results of 360-SS [38], FAOR [40], OmniSSR [33], OSRT [23], and OSRT-light [23] on the ODISR [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Visual comparison of the super-resolution results of 360-SS [38], FAOR [40], OmniSSR [33], OSRT [23], and OSRT-light [23] on the SUN360- [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Visual comparison of the super-resolution results of 360-SS [38], BPOSR [27], FAOR [40], OmniSSR [33], and OSRT [23] on the ODISR [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 9
Figure 9. Figure 9: Visual comparison of the super-resolution results of 360-SS [38], BPOSR [27], FAOR [40], OmniSSR [33], and OSRT [23] on the SUN360- [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]
Figure 10
Figure 10. Figure 10: Visual comparison of the super-resolution results of 360-SS [38], BPOSR [27], FAOR [40], LAU [29], and OmniSSR [33] on the ODISR [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]
Figure 11
Figure 11. Figure 11: Visual comparison of the super-resolution results of 360-SS [38], BPOSR [27], FAOR [40], LAU [29], and OmniSSR [33] on the SUN360- [PITH_FULL_IMAGE:figures/full_fig_p018_11.png]
Figure 12
Figure 12. Figure 12: Visual comparison of the super-resolution results of 360-SS [38], BPOSR [27], and OSRT [23] on the 360Insta dataset at [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

80 extracted references · 77 canonical work pages

  1. [1]

    Low-complexity single-image super-resolution based on nonneg- ative neighbor embedding,

    M. Bevilacqua, A. Roumy, C. Guillemot, and M. L. Alberi-Morel, “Low-complexity single-image super-resolution based on nonneg- ative neighbor embedding,” inBMVC, 2012

    work page 2012

  2. [2]

    Single image super-resolution with non-local means and steering kernel regression,

    K. Zhang, X. Gao, D. Tao, and X. Li, “Single image super-resolution with non-local means and steering kernel regression,”TIP, vol. 21, no. 11, pp. 4544–4556, 2012

    work page 2012

  3. [3]

    Image super-resolution with sparse neighbor embedding,

    X. Gao, K. Zhang, D. Tao, and X. Li, “Image super-resolution with sparse neighbor embedding,”TIP, vol. 21, no. 7, pp. 3194–3205, 2012

    work page 2012

  4. [4]

    Super-resolution from an omnidirectional image sequence,

    H. Nagahara, Y. Yagi, and M. Yachida, “Super-resolution from an omnidirectional image sequence,” inIECON, vol. 4, 2000, pp. 2559–2564

    work page 2000

  5. [5]

    Super-resolution image reconstruction for omni-vision based on pocs,

    F. Zhang and Q. D. Zhu, “Super-resolution image reconstruction for omni-vision based on pocs,” inCCDC, 2009, pp. 5045–5049

    work page 2009

  6. [6]

    Plenoptic based super-resolution for omnidirectional image se- quences,

    L. Bagnato, Y. Boursier, P . Frossard, and P . Vandergheynst, “Plenoptic based super-resolution for omnidirectional image se- quences,” inICIP, 2010, pp. 2829–2832

    work page 2010

  7. [7]

    Joint registration and super-resolution with omnidirectional images,

    Z. Arican and P . Frossard, “Joint registration and super-resolution with omnidirectional images,”TIP, vol. 20, no. 11, pp. 3151–3162, 2011

    work page 2011

  8. [9]

    Deeply-recursive convolutional network for image super-resolution,

    J. Kim, J. K. Lee, and K. M. Lee, “Deeply-recursive convolutional network for image super-resolution,” inCVPR, 2016, pp. 1637– 1645

    work page 2016

Show all 80 references
  1. [10]

    Densely connected convolutional networks,

    G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” inCVPR, 2017, pp. 4700–4708

    2017

  2. [11]

    Memory- friendly scalable super-resolution via rewinding lottery ticket hy- pothesis,

    J. Lin, X. Luo, M. Hong, Y. Qu, Y. Xie, and Z. Wu, “Memory- friendly scalable super-resolution via rewinding lottery ticket hy- pothesis,” inCVPR, 2023, pp. 14 398–14 407

    2023

  3. [12]

    Met- ric learning based interactive modulation for real-world super- resolution,

    C. Mou, Y. Wu, X. Wang, C. Dong, J. Zhang, and Y. Shan, “Met- ric learning based interactive modulation for real-world super- resolution,” inECCV, 2022, pp. 723–740

    2022

  4. [13]

    Content-aware local gan for photo- realistic super-resolution,

    J. Park, S. Son, and K. M. Lee, “Content-aware local gan for photo- realistic super-resolution,” inICCV, 2023, pp. 10 585–10 594

    2023

  5. [14]

    Noise-free optimization in early training steps for image super-resolution,

    M. Lee and J. P . Heo, “Noise-free optimization in early training steps for image super-resolution,” inAAAI, vol. 38, no. 4, 2024, pp. 2920–2928

    2024

  6. [15]

    Srformer: Permuted self-attention for single image super- resolution,

    Y. Zhou, Z. Li, C.-L. Guo, S. Bai, M. M. Cheng, and Q. Hou, “Srformer: Permuted self-attention for single image super- resolution,” inICCV, 2023, pp. 12 780–12 791

    2023

  7. [16]

    Msra- sr: Image super-resolution transformer with multi-scale shared representation acquisition,

    X. Zhou, H. Huang, R. He, Z. Wang, J. Hu, and T. Tan, “Msra- sr: Image super-resolution transformer with multi-scale shared representation acquisition,” inICCV, 2023, pp. 12 665–12 676

    2023

  8. [17]

    Activating more pixels in image super-resolution transformer,

    X. Chen, X. Wang, J. Zhou, Y. Qiao, and C. Dong, “Activating more pixels in image super-resolution transformer,” inCVPR, 2023, pp. 22 367–22 377

    2023

  9. [18]

    Efficient conditional diffusion model with probability flow sampling for image super-resolution,

    Y. Yuan and C. Yuan, “Efficient conditional diffusion model with probability flow sampling for image super-resolution,” inAAAI, vol. 38, no. 7, 2024, pp. 6862–6870

    2024

  10. [19]

    Sinsr: diffusion-based image super-resolution in a single step,

    Y. Wang, W. Yang, X. Chen, Y. Wang, L. Guo, L.-P . Chau, Z. Liu, Y. Qiao, A. C. Kot, and B. Wen, “Sinsr: diffusion-based image super-resolution in a single step,” inCVPR, 2024, pp. 25 796– 25 805

    2024

  11. [20]

    Acdmsr: Accelerated conditional diffusion models for single image super-resolution,

    A. Niu, T. X. Pham, K. Zhang, J. Sun, Y. Zhu, Q. Yan, I. S. Kweon, and Y. Zhang, “Acdmsr: Accelerated conditional diffusion models for single image super-resolution,”TOB, vol. 70, no. 2, pp. 492– 504, 2024

    2024

  12. [21]

    Adadiffsr: Adaptive region-aware dynamic acceleration diffusion model for real-world image super-resolution,

    Y. Fan, C. Liu, N. Yin, C. Gao, and X. Qian, “Adadiffsr: Adaptive region-aware dynamic acceleration diffusion model for real-world image super-resolution,” inECCV, 2025, pp. 396–413

    2025

  13. [22]

    Swinir: Image restoration using swin transformer,

    J. Liang, J. Cao, G. Sun, K. Zhang, L. Van Gool, and R. Timofte, “Swinir: Image restoration using swin transformer,” inICCV, 2021, pp. 1833–1844

    2021

  14. [23]

    Osrt: Om- nidirectional image super-resolution with distortion-aware trans- former,

    F. Yu, X. Wang, M. Cao, G. Li, Y. Shan, and C. Dong, “Osrt: Om- nidirectional image super-resolution with distortion-aware trans- former,” inCVPR, 2023, pp. 13 283–13 292

    2023

  15. [24]

    Perception-oriented omnidirectional image super-resolution based on transformer network,

    H. An and X. Zhang, “Perception-oriented omnidirectional image super-resolution based on transformer network,” inICIP, 2023, pp. 3583–3587

    2023

  16. [25]

    Opdn: Omnidirectional position- aware deformable network for omnidirectional image super- resolution,

    X. Sun, W. Li, Z. Zhang, Q. Ma, X. Sheng, M. Cheng, H. Ma, S. Zhao, J. Zhang, J. Liet al., “Opdn: Omnidirectional position- aware deformable network for omnidirectional image super- resolution,” inCVPR, 2023, pp. 1293–1301

    2023

  17. [26]

    Geometric distortion guided transformer for omnidirec- tional image super-resolution,

    C. Yang, R. Dong, J. Xiao, C. Zhang, K. M. Lam, F. Zhou, and G. Qiu, “Geometric distortion guided transformer for omnidirec- tional image super-resolution,”TCSVT, 2024

    2024

  18. [27]

    Omnidirectional image super-resolution via bi-projection fusion,

    J. Wang, Y. Cui, Y. Li, W. Ren, and X. Cao, “Omnidirectional image super-resolution via bi-projection fusion,” inAAAI, vol. 38, no. 6, 2024, pp. 5454–5462

    2024

  19. [28]

    Spherical pseudo-cylindrical representation for omnidirectional image super-resolution,

    Q. Cai, M. Li, D. Ren, J. Lyu, H. Zheng, J. Dong, and Y. H. Yang, “Spherical pseudo-cylindrical representation for omnidirectional image super-resolution,” inAAAI, vol. 38, no. 2, 2024, pp. 873– 881

    2024

  20. [29]

    Lau-net: Latitude adaptive upscaling network for omnidirectional image super-resolution,

    X. Deng, H. Wang, M. Xu, Y. Guo, Y. Song, and L. Yang, “Lau-net: Latitude adaptive upscaling network for omnidirectional image super-resolution,” inCVPR, 2021, pp. 9189–9198

    2021

  21. [30]

    Omnidirectional image super-resolution via latitude adaptive network,

    X. Deng, H. Wang, M. Xu, L. Li, and Z. Wang, “Omnidirectional image super-resolution via latitude adaptive network,”TMM, vol. 25, pp. 4108–4120, 2022

    2022

  22. [31]

    360 panorama super-resolution using deep convolutional networks,

    V . Fakour-Sevom, E. Guldogan, and J. K. K ¨am¨ar¨ainen, “360 panorama super-resolution using deep convolutional networks,” inVISAPP, vol. 1, 2018, pp. 159–165

    2018

  23. [32]

    Tccl-net: Transformer- convolution collaborative learning network for omnidirectional image super-resolution,

    X. Chai, F. Shao, Q. Jiang, and H. Ying, “Tccl-net: Transformer- convolution collaborative learning network for omnidirectional image super-resolution,”KBS, vol. 274, p. 110625, 2023

    2023

  24. [33]

    Omnissr: Zero-shot omnidi- rectional image super-resolution using stable diffusion model,

    R. Li, X. Sheng, W. Li, and J. Zhang, “Omnissr: Zero-shot omnidi- rectional image super-resolution using stable diffusion model,” in ECCV, 2025, pp. 198–216

    2025

  25. [34]

    Realosr: Latent unfolding boosting diffusion-based real-world omnidirec- tional image super-resolution,

    X. Sheng, R. Li, B. Chen, W. Li, X. Jiang, and J. Zhang, “Realosr: Latent unfolding boosting diffusion-based real-world omnidirec- tional image super-resolution,”arXiv:2412.09646, 2024

    2024

  26. [35]

    Diffosr: Latitude-aware conditional diffusion probabilistic model for omnidirectional image super-resolution,

    L. Liu, T. Luo, G. Jiang, Y. Chen, H. Xu, R. Hu, and Z. He, “Diffosr: Latitude-aware conditional diffusion probabilistic model for omnidirectional image super-resolution,”KBS, vol. 315, p. 113244, 2025

    2025

  27. [36]

    Ntire 2023 challenge on 360deg omni- directional image and video super-resolution: Datasets, methods and results,

    M. Cao, C. Mou, F. Yu, X. Wang, Y. Zheng, J. Zhang, C. Dong, G. Li, Y. Shan, R. Timofteet al., “Ntire 2023 challenge on 360deg omni- directional image and video super-resolution: Datasets, methods and results,” inCVPR, 2023, pp. 1731–1745

    2023

  28. [37]

    360- degree video super resolution and quality enhancement challenge: Methods and results,

    A. Telili, W. Hamidouche, I. Farhat, H. Amirpour, C. Timmerer, I. Khadraoui, J. Lu, T. Van Le, J. Baek, J. Y. Leeet al., “360- degree video super resolution and quality enhancement challenge: Methods and results,”arXiv:2411.06738, 2024

    2024 arXiv

  29. [38]

    Super-resolution of omnidi- rectional images using adversarial learning,

    C. Ozcinar, A. Rana, and A. Smolic, “Super-resolution of omnidi- rectional images using adversarial learning,” inMMSP, 2019, pp. 1–6

    2019

  30. [39]

    Omnidirec- tional video super-resolution using deep learning,

    A. A. Baniya, T. K. Lee, P . W. Eklund, and S. Aryal, “Omnidirec- tional video super-resolution using deep learning,”TMM, vol. 26, pp. 540–554, 2023

    2023

  31. [40]

    Fast omni-directional image super-resolution: Adapting the implicit image function with pixel and semantic-wise spherical geometric priors,

    X. Shen, Y. Wang, S. Zheng, K. Xiao, W. Yang, and X. Wang, “Fast omni-directional image super-resolution: Adapting the implicit image function with pixel and semantic-wise spherical geometric priors,” inAAAI, 2025, pp. 6833–6841

    2025

  32. [41]

    Geometric relationship-guided transformer network for omnidirectional image super-resolution,

    J. Cao, Q. Ding, and H. Luo, “Geometric relationship-guided transformer network for omnidirectional image super-resolution,” SIVP, vol. 19, no. 5, p. 386, 2025. 20

    2025

  33. [42]

    A single frame and multi-frame joint network for 360-degree panorama video super-resolution,

    H. Liu, W. Ma, Z. Ruan, C. Fang, F. Shang, Y. Liu, L. Wang, C. Wang, and D. Jiang, “A single frame and multi-frame joint network for 360-degree panorama video super-resolution,”EAAI, vol. 134, p. 108601, 2024

    2024

  34. [43]

    Spheresr: 360deg image super-resolution with arbitrary projection via continuous spherical image representation,

    Y. Yoon, I. Chung, L. Wang, and K. J. Yoon, “Spheresr: 360deg image super-resolution with arbitrary projection via continuous spherical image representation,” inCVPR, 2022, pp. 5677–5686

    2022

  35. [44]

    Applying vertexshuffle toward 360-degree video super-resolution,

    N. Li and Y. Liu, “Applying vertexshuffle toward 360-degree video super-resolution,” inNOSSDAV, 2022, pp. 71–77

    2022

  36. [45]

    360-degree image super-resolution based on single image sample and progressive residual generative adversarial network,

    L. Qian, X. Liu, J. Wu, X. Xu, and H. Zeng, “360-degree image super-resolution based on single image sample and progressive residual generative adversarial network,” inICIVC, 2022, pp. 654– 661

    2022

  37. [46]

    Fato: Frequency atten- tion transformer for omnidirectional image super-resolution,

    H. An, X. Zhang, S. Zhao, and L. Zhang, “Fato: Frequency atten- tion transformer for omnidirectional image super-resolution,” in MMASIA, 2024, pp. 1–7

    2024

  38. [47]

    A generic camera model and calibra- tion method for conventional, wide-angle, and fish-eye lenses,

    J. Kannala and S. S. Brandt, “A generic camera model and calibra- tion method for conventional, wide-angle, and fish-eye lenses,” TP AMI, vol. 28, no. 8, pp. 1335–1340, 2006

    2006

  39. [48]

    Survey on fish-eye cameras and their applications in intelligent vehicles,

    Y. Qian, M. Yang, and J. M. Dolan, “Survey on fish-eye cameras and their applications in intelligent vehicles,”TITS, vol. 23, no. 12, pp. 22 755–22 771, 2022

    2022

  40. [49]

    360 single image super resolution via distortion-aware network and distorted perspective images,

    A. Nishiyama, S. Ikehata, and K. Aizawa, “360 single image super resolution via distortion-aware network and distorted perspective images,” inICIP, 2021, pp. 1829–1833

    2021

  41. [50]

    An omnidirectional image super- resolution method based on enhanced swinir,

    X. Yao, Y. Pan, and J. Wang, “An omnidirectional image super- resolution method based on enhanced swinir,”Information, vol. 15, no. 5, p. 248, 2024

    2024

  42. [51]

    Omnidirectional image super-resolution via position attention network,

    X. Wang, S. Wang, J. Li, M. Li, J. Li, and Y. Xu, “Omnidirectional image super-resolution via position attention network,”NN, vol. 178, p. 106464, 2024

    2024

  43. [52]

    Mambaosr: Leveraging spatial- frequency mamba for distortion-guided omnidirectional image super-resolution,

    W. Wen, Q. Zhao, and X. Shao, “Mambaosr: Leveraging spatial- frequency mamba for distortion-guided omnidirectional image super-resolution,”Entropy, vol. 27, no. 4, p. 446, 2025

    2025

  44. [53]

    Image super-resolution using deep convolutional networks,

    C. Dong, C. C. Loy, K. He, and X. Tang, “Image super-resolution using deep convolutional networks,”TP AMI, vol. 38, no. 2, pp. 295–307, 2015

    2015

  45. [54]

    Weighted-to-spherically-uniform quality evaluation for omnidirectional video,

    Y. Sun, A. Lu, and L. Yu, “Weighted-to-spherically-uniform quality evaluation for omnidirectional video,”SPL, vol. 24, no. 9, pp. 1408–1412, 2017

    2017

  46. [55]

    Attention is all you need,

    A. Vaswani, “Attention is all you need,”NIPS, vol. 36, 2017

    2017

  47. [56]

    Bifuse++: Self-supervised and efficient bi-projection fusion for 360 depth estimation,

    F. E. Wang, Y. H. Yeh, Y. H. Tsai, W. C. Chiu, and M. Sun, “Bifuse++: Self-supervised and efficient bi-projection fusion for 360 depth estimation,”TP AMI, vol. 45, no. 5, pp. 5448–5460, 2022

    2022

  48. [57]

    Spatio-temporal distortion aware omnidirectional video super-resolution,

    H. An, X. Zhang, L. Zhang, and R. Xiong, “Spatio-temporal distortion aware omnidirectional video super-resolution,” arXiv:2410.11506, 2024

    2024 arXiv

  49. [58]

    Basicvsr++: Im- proving video super-resolution with enhanced propagation and alignment,

    K. C. Chan, S. Zhou, X. Xu, and C. C. Loy, “Basicvsr++: Im- proving video super-resolution with enhanced propagation and alignment,” inCVPR, 2022, pp. 5972–5981

    2022

  50. [59]

    Modeling the perceptual quality of immersive images rendered on head mounted displays: Resolution and compres- sion,

    M. Huang, Q. Shen, Z. Ma, A. C. Bovik, P . Gupta, R. Zhou, and X. Cao, “Modeling the perceptual quality of immersive images rendered on head mounted displays: Resolution and compres- sion,”TIP, vol. 27, no. 12, pp. 6039–6050, 2018

    2018

  51. [60]

    Recognizing scene viewpoint using panoramic place representation,

    J. Xiao, K. A. Ehinger, A. Oliva, and A. Torralba, “Recognizing scene viewpoint using panoramic place representation,” inCVPR, 2012, pp. 2695–2702

    2012

  52. [61]

    Bridge the gap between vqa and human behavior on omnidirectional video: A large-scale dataset and a deep learning model,

    C. Li, M. Xu, X. Du, and Z. Wang, “Bridge the gap between vqa and human behavior on omnidirectional video: A large-scale dataset and a deep learning model,” inACMM, 2018, pp. 932–940

    2018

  53. [62]

    Learning compressible 360 video isomers,

    Y.-C. Su and K. Grauman, “Learning compressible 360 video isomers,” inCVPR, 2018, pp. 7824–7833

    2018

  54. [63]

    A saliency dataset for 360-degree videos,

    A. Nguyen and Z. Yan, “A saliency dataset for 360-degree videos,” inMMSys, 2019, pp. 279–284

    2019

  55. [64]

    A survey on 360-degree video: Coding, quality of experience and streaming,

    F. Chiariotti, “A survey on 360-degree video: Coding, quality of experience and streaming,”Comput. Commun., vol. 177, pp. 133– 155, 2021

    2021

  56. [65]

    State-of-the-art in 360 video/image processing: Perception, assessment and compres- sion,

    M. Xu, C. Li, S. Zhang, and P . Le Callet, “State-of-the-art in 360 video/image processing: Perception, assessment and compres- sion,”JSTSP, vol. 14, no. 1, pp. 5–26, 2020

    2020

  57. [66]

    A taxonomy and dataset for 360 videos,

    A. T. Nasrabadi, A. Samiei, A. Mahzari, R. P . McMahan, R. Prakash, M. C. Farias, and M. M. Carvalho, “A taxonomy and dataset for 360 videos,” inMMSys, 2019, pp. 273–278

    2019

  58. [67]

    Personal identifiability of user tracking data during observation of 360-degree vr video,

    M. R. Miller, F. Herrera, H. Jun, J. A. Landay, and J. N. Bailenson, “Personal identifiability of user tracking data during observation of 360-degree vr video,”Sci. Rep., vol. 10, no. 1, p. 17404, 2020

    2020

  59. [68]

    Rethinking alignment in video super-resolution transformers,

    S. Shi, J. Gu, L. Xie, X. Wang, Y. Yang, and C. Dong, “Rethinking alignment in video super-resolution transformers,”NIPS, vol. 35, pp. 36 081–36 093, 2022

    2022

  60. [69]

    Enhancing video super-resolution via implicit resampling-based alignment,

    K. Xu, Z. Yu, X. Wang, M. B. Mi, and A. Yao, “Enhancing video super-resolution via implicit resampling-based alignment,” inCVPR, 2024, pp. 2546–2555

    2024

  61. [70]

    Video super-resolution transformer with masked inter&intra-frame at- tention,

    X. Zhou, L. Zhang, X. Zhao, K. Wang, L. Li, and S. Gu, “Video super-resolution transformer with masked inter&intra-frame at- tention,” inCVPR, 2024, pp. 25 399–25 408

    2024

  62. [71]

    Investigating tradeoffs in real-world video super-resolution,

    K. C. Chan, S. Zhou, X. Xu, and C. C. Loy, “Investigating tradeoffs in real-world video super-resolution,” inCVPR, 2022, pp. 5962– 5971

    2022

  63. [72]

    Savsr: arbitrary-scale video super-resolution via a learned scale-adaptive network,

    Z. Li, H. Liu, F. Shang, Y. Liu, L. Wan, and W. Feng, “Savsr: arbitrary-scale video super-resolution via a learned scale-adaptive network,” inAAAI, vol. 38, no. 4, 2024, pp. 3288–3296

    2024

  64. [73]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P . Simoncelli, “Image quality assessment: from error visibility to structural similarity,” TIP, vol. 13, no. 4, pp. 600–612, 2004

    2004

  65. [74]

    Weighted- to-spherically-uniform ssim objective quality evaluation for panoramic video,

    Y. Zhou, M. Yu, H. Ma, H. Shao, and G. Jiang, “Weighted- to-spherically-uniform ssim objective quality evaluation for panoramic video,” inICSP, 2018, pp. 54–57

    2018

  66. [75]

    Making a “com- pletely blind

    A. Mittal, R. Soundararajan, and A. C. Bovik, “Making a “com- pletely blind” image quality analyzer,”SPL, vol. 20, no. 3, pp. 209–212, 2012

    2012

  67. [76]

    Musiq: Multi- scale image quality transformer,

    J. Ke, Q. Wang, Y. Wang, P . Milanfar, and F. Yang, “Musiq: Multi- scale image quality transformer,” inICCV, 2021, pp. 5148–5157

    2021

  68. [77]

    Maniqa: Multi-dimension attention network for no- reference image quality assessment,

    S. Yang, T. Wu, S. Shi, S. Lao, Y. Gong, M. Cao, J. Wang, and Y. Yang, “Maniqa: Multi-dimension attention network for no- reference image quality assessment,” inCVPR, 2022, pp. 1191– 1200

    2022

  69. [78]

    Exploring clip for assessing the look and feel of images,

    J. Wang, K. C. Chan, and C. C. Loy, “Exploring clip for assessing the look and feel of images,” inAAAI, vol. 37, no. 2, 2023, pp. 2555–2563

    2023

  70. [79]

    Sora: A review on background, technology, limitations, and opportunities of large vision models,

    Y. Liu, K. Zhang, Y. Li, Z. Yan, C. Gao, R. Chen, Z. Yuan, Y. Huang, H. Sun, J. Gao, L. He, and L. Sun, “Sora: A review on background, technology, limitations, and opportunities of large vision models,” arXiv:2402.17177, 2024

    2024 arXiv

  71. [80]

    Seesr: Towards semantics-aware real-world image super-resolution,

    R. Wu, T. Yang, L. Sun, Z. Zhang, S. Li, and L. Zhang, “Seesr: Towards semantics-aware real-world image super-resolution,” in CVPR, 2024, pp. 25 456–25 467

    2024

  72. [81]

    Upscale-a-video: Temporal-consistent diffusion model for real-world video super- resolution,

    S. Zhou, P . Yang, J. Wang, Y. Luo, and C. C. Loy, “Upscale-a-video: Temporal-consistent diffusion model for real-world video super- resolution,” inCVPR, 2024, pp. 2535–2545

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.