Pith. sign in

REVIEW 4 major objections 6 minor 34 references

MIORe & VAR-MIORe: Benchmarks to Push the Boundaries of Restoration

T0 review · 4 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read Two 1000-FPS datasets benchmark restoration from sharp to extreme motion

desk verdict A potentially useful multi-task benchmark with a real gap to fill, but the central blur-synthesis rule is under-specified and the OF ground truth is circular; needs a release and a precise formula before it can be trusted. read the letter →

arxiv 2509.06803 v1 pith:5T6NRRLQ submitted 2025-09-08 cs.CV

George Ciubotariu , Zhuyun Zhou , Zongwei Wu , Radu Timofte This is my paper · ORCID
classification cs.CV
keywords motiondeblurringvideoframeinterpolationopticalflowestimationhigh-frame-ratedatasetadaptiveaveragingmagnitudecontrolmulti-taskbenchmark
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper introduces MIORe and VAR-MIORe, two multi-task datasets designed to overcome what the authors see as critical gaps in motion-restoration benchmarks: fixed or limited motion magnitudes, lack of controlled blur, and no unified ground truth for deblurring, video frame interpolation, and optical flow. Recorded at 1000 FPS with professional lenses, the datasets generate blurry frames by adaptively averaging a variable number of sharp frames, with the middle sharp frame kept as ground truth. VAR-MIORe extends the range to extreme motion (maximum optical flow of 1932 pixels) and is claimed to be the first benchmark with explicit control over motion amplitude. If these claims hold, researchers gain a single testbed that can reveal where restoration algorithms break as motion intensity grows.

What carries the argument

Adaptive frame averaging driven by the mean and maximum optical flow computed with DEQFlow: fast scenes (about 10 pixels of flow per frame) are averaged over few frames (around 3), while subpixel scenes are averaged over many frames (up to 30), targeting a mid-level optical flow of about 30 pixels per blurry image. This mechanism controls blur magnitude while keeping the left, middle, and right sharp frames as ground truth for VFI, optical flow, and deblurring.

What would settle it

Apply the paper's adaptive averaging to synthetic sequences with known true optical flow and measure the resulting blur's flow; a large spread around the claimed 30-pixel target would refute the consistency claim. Separately, recompute optical flow on the sharp frame pairs with a method independent of DEQFlow and compare with the pseudo-labels; systematic deviation would expose the circularity of the optical-flow ground truth.

Watch

Extended reading notes

Core claim

The central claim is that high-frame-rate acquisition combined with flow-based adaptive frame averaging yields consistent, controlled motion blur while preserving completely sharp flanking and middle frames, enabling the same sequences to serve as ground truth for motion deblurring, video frame interpolation, and optical flow estimation. VAR-MIORe is further claimed to provide the first explicit control over motion amplitude by spanning from static to extreme motion, allowing systematic study of model 'breaking points.' Benchmark results on state-of-the-art methods trained on existing datasets show measurable performance drops across the new datasets, with frequency-domain deblurring methods

Load-bearing premise

The rule that chooses how many sharp frames to average into each blurry frame is described only by examples (10 pixels per frame yields 3 frames; subpixel motion yields up to 30) and never as an exact function, so the promised consistent and controlled blur level rests on an ad hoc, underspecified step.

Editorial extensions

If this is right

  • A single dataset now supplies aligned inputs for deblurring, video frame interpolation, and optical flow, enabling multi-task training and evaluation on identical frames.
  • VAR-MIORe's continuous motion-magnitude axis lets researchers plot performance versus blur strength and identify the regime where each method breaks.
  • The reported drop of some deblurring methods on mild motion suggests that training exclusively on fast-motion benchmarks may bias models toward high-frequency priors that misbehave on slow scenes.
  • Including defocus, exposure, and weather conditions pushes restoration algorithms to handle intertwined optical degradations rather than clean motion only.

Reading between the lines

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

  • Because the same DEQFlow method generates the blur (via averaging decisions) and the optical-flow pseudo-labels, the OF evaluation is partly circular; an independent flow estimator would likely change the reported rankings.
  • The adaptive averaging rule is underspecified — only examples are given, not an exact formula — so reproducibility and the claimed 30-pixel blur target would need a precise mapping to verify.
  • The 'variable motion amplitude' design could become a standard stress test for restoration models, analogous to rate-distortion sweeps in compression, giving a one-dimensional axis along which to compare robustness.
  • The observation that frequency-prior methods hallucinate in slow motion implies that benchmarks should deliberately include low-motion regimes, which dominate real consumer footage yet are filtered out by many existing datasets.
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 / 6 minor

Summary. This paper introduces two multi-task video datasets, MIORe and VAR-MIORe, built from 1000 FPS 1920x1080 recordings of 333 in-the-wild sequences captured with professional prime/zoom lenses. Blurry images are synthesized by averaging varying numbers of high-speed frames, with the number of frames selected from computed optical-flow statistics; the left, middle, and right sharp frames are kept as inputs/ground truth for video frame interpolation, motion deblurring, and optical flow estimation. VAR-MIORe extends MIORe to a continuous range of integration offsets (1-249 ms) to explicitly vary motion amplitude. The paper benchmarks several state-of-the-art deblurring, VFI, and OF methods and reports results by motion-intensity splits and integration buckets.

Significance. If the construction is made fully specified, the datasets could be a genuinely useful resource: the 1000 FPS acquisition, professional optics, diverse weather/seasonal conditions, multi-task paired sharp/blurry data, and motion-intensity buckets address real limitations of existing benchmarks such as GoPro, Vimeo90K, and Sintel. The reported benchmarking also gives an initial picture of how SOTA methods behave across motion amplitudes. The paper is transparent that the OF labels are pseudo-ground-truth and that the OF results are preliminary. However, the central adaptive-averaging rule is not fully specified, and the OF pseudo-labeling is entangled with the blur-generation procedure, so the claims of consistent blur control and precise multi-task ground truth are not yet established.

major comments (4)
  1. [Sec. 4.3, Fig. 8] The adaptive frame-averaging rule is not specified as a function. The text states a target of 30 pixels of optical flow and gives examples: about 10 px/frame leads to 3 averaged frames, and subpixel motion leads to up to 30 frames. If total blur magnitude is approximately per-frame flow times the number of averaged frames, then a scene with v=0.5 px/frame would require N=60 frames to reach the 30-pixel target, but the stated cap is 30, so the resulting blur would be only about 15 px. Thus the claimed 'consistent blur level' and 'homogeneity' are not guaranteed by the described procedure. The six bins in Fig. 8 and the role of the 'mean and maximum optical flow' mentioned in Sec. 4.3 are also not formally defined. Since frame averaging is the core dataset-construction mechanism, this is a reproducibility and correctness blocker. Please provide the exact mapping, pseudocode, thresholds, an
  2. [Sec. 4.3 / Sec. 5.3] The optical flow benchmark is circular: the pseudo-ground-truth flow labels are generated with DEQFlow [2], which is the same method used in Sec. 4.3 to compute optical flow for selecting the frame-averaging counts. Consequently, the OF evaluation measures agreement with the estimator that defines the blur intensity, and systematic errors in DEQFlow are baked into both the labels and the blur synthesis. The paper should explicitly state this dependence in Sec. 4.3 as well as in Sec. 5.3, and should provide an independent validation subset (e.g., synthetic sequences with known flow, or labels from an independent estimator) to calibrate the pseudo-labels before claiming 'precise ground truth' for optical flow.
  3. [Sec. 4.4, Table 3] The reported dataset statistics do not reconcile with the bucket description. Table 3 says VAR-MIORe is organized in buckets from 1 to 249 with 333 input entries per bucket; 249 x 333 = 82,917, not the stated 83,250. It is also unclear whether the seven displayed columns (1, 5, 13, 29, 61, 125, 249) are a subsample of 249 buckets or the only buckets. The 52,218-frame size of MIORe and the per-split counts also cannot be derived from the 333 sequences and the sequence counts in Fig. 8. Please report exact per-bucket counts, the full set of frame-count values, split assignments, and the total number of blurry images for each dataset.
  4. [Table 6] The Total column for VideoFlow is internally inconsistent. Using the split sizes given in Fig. 8 (119 Extreme, 88 Fast, 35 Medium-Fast, 38 Medium, 35 Mild-Medium, 18 Mild), the weighted average of the six EPE values 6.17, 8.44, 9.12, 7.69, 9.57, 8.88 is approximately 7.76, not 19.38. The corresponding F1 total (63.74) does equal the weighted average, and the Total values for the other four methods also behave as weighted averages. Unless a different split or weighting is intended, this appears to be a reporting error in a headline result. Please correct it and state the exact split sizes and weighting used for Total columns.
minor comments (6)
  1. [Table 5] The PerVFI row cites reference [15], but PerVFI is reference [28] in the bibliography and in Table 4. Please fix the citation.
  2. [Fig. 8] The figure layout is very hard to read after extraction: the mapping between the six category labels and the two numeric rows (Optical Flow Magnitude and # Sharp Frames) is ambiguous. Provide a clean tabular legend or a separate table with exact ranges and counts.
  3. [Sec. 3.3] The claim that MIORe and VAR-MIORe are 'the first in the motion deblurring literature to integrate defocus blur as an intrinsic component' is strong and should be supported by a comparison with existing datasets that combine defocus and motion blur, or by a focused literature search.
  4. [Sec. 5.3] The F1 metric for optical flow is not defined. Since the labels are pseudo-ground-truth, please specify the threshold used for F1 and how the EPE is computed (e.g., on which regions, ignoring invalid/occluded pixels).
  5. [Sec. 4.1 / Table 1] The phrase 'sustaining 1000 FPS for both acquisition and final take-away dataset' conflicts with the Table 1 entries 'FPS (Acq. / Final)' of [28,1000] for MIORe and [4,1000] for VAR-MIORe. Clarify whether the final dataset frame rate is the raw 1000 FPS sharp frames or the effective rate of blurry frames.
  6. [Sec. 4.2] Minor wording issue: 'successfully ensuring consistent framing and no motion artifacts' is a run-on; also, 'for several sequences, we use a tripod... successfully ensuring' should be rephrased.

Circularity Check

1 steps flagged · score 4.0 of 10

Partial circularity in optical-flow pseudo-labels: DEQFlow both controls blur synthesis and supplies OF ground truth.

  1. fitted input called prediction [Sec 4.3 (Dataset Construction) and Sec 5.3 (Optical Flow Estimation)]
    "For each sequence, we determine the optimal blur intensity by computing both the mean and maximum optical flow. These metrics inform the selection of the number of frames to average when generating blurry images ... We adopt established optical flow computation methodologies from prior works [3, 24] and enhance them with the latest approaches [2]. ... Table 6 compares the performance of five methods evaluated on pseudo-ground truth labels generated following the approach in [2]."

    The same DEQFlow estimator [2] is used twice: in Sec 4.3 its mean/max optical flow determines how many sharp frames are averaged to create each blurry image, and in Sec 5.3 its output is the optical-flow 'ground truth' against which methods are scored. The target of the OF benchmark is therefore the output of the same estimator that calibrated the test-input blur, not an independently measured motion field. Evaluating other flow methods against this target measures agreement with DEQFlow rather than physical motion, so the OF evaluation is forced by the construction. The paper labels these labels 'pseudo-ground truth' and calls the results a proof of concept, which is honest but does not remove the dependency. Deblurring and VFI use real sharp frames, so their benchmarks are not affected.

full rationale

There is no self-citation chain, no imported uniqueness theorem, and no ansatz smuggled in via citation. The only load-bearing circular dependency is the dual use of DEQFlow: the same optical-flow estimator is used to select the frame-averaging count for blur synthesis (Sec 4.3) and to generate the optical-flow pseudo-ground-truth labels (Sec 5.3). This makes the OF evaluation partially self-referential, but the paper openly states that these are pseudo-labels requiring further refinement and presents OF results only as a proof of concept. The deblurring and video-frame-interpolation benchmarks rely on real captured sharp frames as ground truth, so the central multi-task dataset claim has independent content. The adaptive frame-averaging rule is underspecified and potentially inconsistent for very slow scenes, but that is a reproducibility/correctness concern rather than a circularity. Overall, the circularity is real but confined to the OF task, giving a score of 4 rather than higher.

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

The main free parameters are dataset construction choices rather than fitted model constants. The adaptive averaging formula is not specified, and the OF pseudo-labels are both a construction input and the evaluation target.

free parameters (3)
  • target optical flow magnitude = 30 pixels
    Chosen as the mid-level benchmark for synthesized blur in Sec 4.3; the paper does not justify this value.
  • motion intensity bin thresholds = MILD, MILD-MEDIUM, MEDIUM, MEDIUM-FAST, FAST, EXTREME (Figure 8)
    Bins used to split MIORe into XF, F, MF, M, MS, S; boundaries are set by the authors without statistical justification.
  • frames per blurry image = 3 to 35 for MIORe, 1 to 249 for VAR-MIORe
    The number of averaged frames per sequence is the main design variable; the exact adaptive formula is not given.
assumptions (3)
  • domain assumption High-speed 1000 FPS capture preserves true scene motion without temporal aliasing.
    Sec 4.2; the whole pipeline assumes the Chronos 2.1-HD camera at 1000 FPS is accurate and rolling-shutter-free.
  • domain assumption Optical flow pseudo-labels generated by the DEQFlow method [2] are accurate enough to guide blur synthesis and serve as ground truth.
    Sec 4.3 and 5.3; this is the circular step, since the same algorithm is used for synthesis and evaluation.
  • standard math Averaging consecutive sharp frames produces realistic motion blur.
    Sec 4; standard GoPro-style synthesis, but the paper adds no validation that averaged blur matches real camera blur.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MIORe & VAR-MIORe: Benchmarks to Push the Boundaries of Restoration." pith.science (2026). https://pith.science/paper/5T6NRRLQ

@misc{pith2026250906803,
  author       = {Pith},
  title        = {Pith review of: MIORe & VAR-MIORe: Benchmarks to Push the Boundaries of Restoration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5T6NRRLQ}},
  note         = {Machine review of arXiv:2509.06803}
}
read the original abstract

We introduce MIORe and VAR-MIORe, two novel multi-task datasets that address critical limitations in current motion restoration benchmarks. Designed with high-frame-rate (1000 FPS) acquisition and professional-grade optics, our datasets capture a broad spectrum of motion scenarios, which include complex ego-camera movements, dynamic multi-subject interactions, and depth-dependent blur effects. By adaptively averaging frames based on computed optical flow metrics, MIORe generates consistent motion blur, and preserves sharp inputs for video frame interpolation and optical flow estimation. VAR-MIORe further extends by spanning a variable range of motion magnitudes, from minimal to extreme, establishing the first benchmark to offer explicit control over motion amplitude. We provide high-resolution, scalable ground truths that challenge existing algorithms under both controlled and adverse conditions, paving the way for next-generation research of various image and video restoration tasks.

Figures

Figures reproduced from arXiv: 2509.06803 by the authors.

Figure 1
Figure 1. Continuous Motion Magnitude Range Captured by Our Novel Datasets. The figure illustrates some sequences of represen￾tative frames spanning a wide dynamic range of motion magnitude (from sharp to extreme). Numerical annotations denote the integration window parameters: for instance, “Sharp” indicates a minimal integration span of only “1” frame, namely “#125” yielding a sharp image, while “Extreme” corresponds to max… view at source ↗
Figure 2
Figure 2. Visualization of a panning shot from our proposed [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Variety of Adverse Conditions. The figure displays representative examples of challenging environmental conditions encountered in real-world imaging: Clouds, Fog, Rain, and Snow. These conditions introduce distinct degradations, ranging from re￾duced contrast and visibility under fog and clouds to dynamic, transient artifacts in rain and snow, that pose significant challenges for tasks such as deblurring, segmentati… view at source ↗
Figures from the paper (3 more)
Figure 6
Figure 6. Figure 6: Dataset sequence annotations statistics. Our datasets contain samples exhibiting diverse motion patterns, including both subject motion and camera motion, characterized by translations along the Ox, Oy, and Oz axis, as well as rotations in Yaw, Pitch, and Roll. These s…
Figure 7
Figure 7. Figure 7: Diverse Motion Patterns Captured by Our Proposed Datasets. The figure presents a wide spectrum of motion types observed in our dataset. The top row shows samples of blurred images with ego-camera induced motions include Zooming, Tilt Shot, Barrel Roll, and Panning; and…
Figure 8
Figure 8. Figure 8: Motion Magnitude Distribution and Sharp-to-Blurry Frame Ratios. We categorize sequences into six motion inten￾sity groups, ranging from 18 MILD to 119 EXTREME and based on average optical flow magnitude, which increases progressively across categories. As motion intens…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 33 canonical work pages

  1. [2]

    Zico Kolter

    Shaojie Bai, Zhengyang Geng, Yash Savani, and J. Zico Kolter. Deep equilibrium optical flow estimation. InCVPR. IEEE, 2022. 6, 8

    work page 2022

  2. [1]

    Brown, Jihwan Choe, and Kihwan Kim

    Abdelrahman Abdelhamed, Jonghwa Yim, Abhijith Punnap- purath, Michael S. Brown, Jihwan Choe, and Kihwan Kim. Extracting vignetting and grain filter effects from photos. In WACV. IEEE, 2022. 5

    work page 2022

  3. [3]

    Butler, Jonas Wulff, Garrett B

    Daniel J. Butler, Jonas Wulff, Garrett B. Stanley, and Michael J. Black. A naturalistic open source movie for opti- cal flow evaluation. InECCV. Springer, 2012. 1, 3, 5, 6

    work page 2012

  4. [4]

    Simple baselines for image restoration

    Liangyu Chen, Xiaojie Chu, Xiangyu Zhang, and Jian Sun. Simple baselines for image restoration. InECCV. Springer,

  5. [5]

    Memflow: Optical flow esti- mation and prediction with memory

    Qiaole Dong and Yanwei Fu. Memflow: Optical flow esti- mation and prediction with memory. InCVPR. IEEE, 2024. 3, 8

    work page 2024

  6. [6]

    Flownet: Learn- ing optical flow with convolutional networks

    Alexey Dosovitskiy, Philipp Fischer, Eddy Ilg, Philip H¨ausser, Caner Hazirbas, Vladimir Golkov, Patrick van der Smagt, Daniel Cremers, and Thomas Brox. Flownet: Learn- ing optical flow with convolutional networks. InICCV. IEEE, 2015. 1, 3

    work page 2015

  7. [7]

    Rethinking RAFT for Efficient Optical Flow

    Navid Eslami, Farnoosh Arefi, Amir M. Mansourian, and Shohreh Kasaei. Rethinking RAFT for efficient optical flow. CoRR, abs/2401.00833, 2024. 3, 8

    work page Pith review arXiv 2024

  8. [8]

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

    Zhenxuan Fang, Fangfang Wu, Weisheng Dong, Xin Li, Jin- jian Wu, and Guangming Shi. Self-supervised non-uniform kernel estimation with flow-based motion prior for blind im- age deblurring. InCVPR. IEEE, 2023. 3, 7, 8

    work page 2023

Show all 34 references
  1. [9]

    Are we ready for autonomous driving? the KITTI vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the KITTI vision benchmark suite. InCVPR. IEEE, 2012. 5

    2012

  2. [10]

    Are we ready for autonomous driving? the KITTI vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the KITTI vision benchmark suite. InCVPR. IEEE, 2012. 2, 3, 5, 6

    2012

  3. [11]

    Flowformer: A transformer architecture for optical flow

    Zhaoyang Huang, Xiaoyu Shi, Chao Zhang, Qiang Wang, Ka Chun Cheung, Hongwei Qin, Jifeng Dai, and Hongsheng Li. Flowformer: A transformer architecture for optical flow. InECCV. Springer, 2022. 3, 8

    2022

  4. [12]

    Automatic removal of chromatic aberration from a single image

    Sing Bing Kang. Automatic removal of chromatic aberration from a single image. InCVPR. IEEE, 2007. 5

    2007

  5. [13]

    Mohler, Bernhard Sch¨olkopf, and Stefan Harmeling

    Rolf K ¨ohler, Michael Hirsch, Betty J. Mohler, Bernhard Sch¨olkopf, and Stefan Harmeling. Recording and playback of camera shake: Benchmarking blind deconvolution with a real-world database. InECCV. Springer, 2012. 2

    2012

  6. [14]

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

    Lingshun Kong, Jiangxin Dong, Jianjun Ge, Mingqiang Li, and Jinshan Pan. Efficient frequency domain-based trans- formers for high-quality image deblurring. InCVPR. IEEE,

  7. [15]

    Sparse global matching for video frame interpolation with large motion

    Chunxu Liu, Guozhen Zhang, Rui Zhao, and Limin Wang. Sparse global matching for video frame interpolation with large motion. InCVPR. IEEE, 2024. 3, 8

    2024

  8. [16]

    Flowdiffuser: Advancing optical flow es- timation with diffusion models

    Ao Luo, Xin Li, Fan Yang, Jiangyu Liu, Haoqiang Fan, and Shuaicheng Liu. Flowdiffuser: Advancing optical flow es- timation with diffusion models. InCVPR. IEEE, 2024. 3, 8

    2024

  9. [17]

    Adarevd: Adaptive patch exiting reversible decoder pushes the limit of image deblurring

    Xintian Mao, Qingli Li, and Yan Wang. Adarevd: Adaptive patch exiting reversible decoder pushes the limit of image deblurring. InCVPR. IEEE, 2024. 3, 7, 8

    2024

  10. [18]

    Loformer: Local frequency transformer for im- age deblurring

    Xintian Mao, Jiansheng Wang, Xingran Xie, Qingli Li, and Yan Wang. Loformer: Local frequency transformer for im- age deblurring. InMM. ACM, 2024. 3, 7, 8

    2024

  11. [19]

    A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation

    Nikolaus Mayer, Eddy Ilg, Philip H ¨ausser, Philipp Fischer, Daniel Cremers, Alexey Dosovitskiy, and Thomas Brox. A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation. InCVPR. IEEE,

  12. [20]

    Deep multi-scale convolutional neural network for dynamic scene deblurring

    Seungjun Nah, Tae Hyun Kim, and Kyoung Mu Lee. Deep multi-scale convolutional neural network for dynamic scene deblurring. InCVPR. IEEE, 2017. 1, 3, 5, 6, 7

    2017

  13. [21]

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

    Jaesung Rim, Haeyun Lee, Jucheol Won, and Sunghyun Cho. Real-world blur dataset for learning and benchmarking de- blurring algorithms. InECCV. Springer, 2020. 3, 5, 6

    2020

  14. [22]

    Human-aware mo- tion deblurring

    Ziyi Shen, Wenguan Wang, Xiankai Lu, Jianbing Shen, Haibin Ling, Tingfa Xu, and Ling Shao. Human-aware mo- tion deblurring. InICCV. IEEE, 2019. 5

    2019

  15. [23]

    Videoflow: Exploiting temporal cues for multi-frame optical flow estimation

    Xiaoyu Shi, Zhaoyang Huang, Weikang Bian, Dasong Li, Manyuan Zhang, Ka Chun Cheung, Simon See, Hongwei Qin, Jifeng Dai, and Hongsheng Li. Videoflow: Exploiting temporal cues for multi-frame optical flow estimation. In ICCV. IEEE, 2023. 3, 8

    2023

  16. [24]

    XVFI: ex- treme video frame interpolation

    Hyeonjun Sim, Jihyong Oh, and Munchurl Kim. XVFI: ex- treme video frame interpolation. InICCV. IEEE, 2021. 3, 5, 6

    2021

  17. [25]

    Deep video deblurring for hand-held cameras

    Shuochen Su, Mauricio Delbracio, Jue Wang, Guillermo Sapiro, Wolfgang Heidrich, and Oliver Wang. Deep video deblurring for hand-held cameras. InCVPR. IEEE, 2017. 1, 5

    2017

  18. [26]

    Edge-based blur kernel estimation using patch priors

    Libin Sun, Sunghyun Cho, Jue Wang, and James Hays. Edge-based blur kernel estimation using patch priors. In ICCP. IEEE, 2013. 2

    2013

  19. [27]

    Chronos 2.1-hd high-speed camera

    Kron Technologies. Chronos 2.1-hd high-speed camera. https://www.krontech.ca/product/chronos- 2- 1- hd- high- speed- camera/, 2024. Accessed: 2024-11-11. 5

    2024

  20. [28]

    Perception-oriented video frame interpolation via asymmetric blending

    Guangyang Wu, Xin Tao, Changlin Li, Wenyi Wang, Xiao- hong Liu, and Qingqing Zheng. Perception-oriented video frame interpolation via asymmetric blending. InCVPR. IEEE, 2024. 3, 8

    2024

  21. [29]

    Tianfan Xue, Baian Chen, Jiajun Wu, Donglai Wei, and William T. Freeman. Video enhancement with task-oriented flow.IJCV, 127(8):1106–1125, 2019. 2, 3, 5, 6

    2019

  22. [30]

    Zongsheng Yue, Qian Zhao, Jianwen Xie, Lei Zhang, Deyu Meng, and Kwan-Yee K. Wong. Blind image super- resolution with elaborate degradation modeling on noise and kernel. InCVPR. IEEE, 2022. 5

    2022

  23. [31]

    Extracting motion and ap- pearance via inter-frame attention for efficient video frame interpolation

    Guozhen Zhang, Yuhan Zhu, Haonan Wang, Youxin Chen, Gangshan Wu, and Limin Wang. Extracting motion and ap- pearance via inter-frame attention for efficient video frame interpolation. InCVPR. IEEE, 2023. 3, 8

    2023

  24. [32]

    Vfimamba: Video frame interpo- lation with state space models

    Guozhen Zhang, Chunxu Liu, Yutao Cui, Xiaotong Zhao, Kai Ma, and Limin Wang. Vfimamba: Video frame interpo- lation with state space models. InNIPS, 2024. 3, 8 9

    2024

  25. [33]

    A closer look at blind super-resolution: Degradation models, baselines, and performance upper bounds

    Wenlong Zhang, Guangyuan Shi, Yihao Liu, Chao Dong, and Xiao-Ming Wu. A closer look at blind super-resolution: Degradation models, baselines, and performance upper bounds. InCVPRW. IEEE, 2022. 5

    2022

  26. [34]

    Lei Zhu, Xinjiang Wang, Zhanghan Ke, Wayne Zhang, and Rynson W. H. Lau. Biformer: Vision transformer with bi- level routing attention. InCVPR. IEEE, 2023. 3, 8 10

    2023

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.