Pith. sign in

REVIEW 4 major objections 5 minor 39 references

Elevating Flow-Guided Video Inpainting with Reference Generation

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper claims that video inpainting improves by decoupling known-pixel propagation from generative filling, and reports the best object-removal scores on a new 2K benchmark.

desk verdict A solid, well-engineered flow-based video inpainting system with a genuinely new one-shot pulling mechanism and a useful benchmark; the evaluation is good but overclaims significance and leaves the flow-completion dependency untested. read the letter →

arxiv 2412.08975 v1 pith:PBAQMRNO submitted 2024-12-12 cs.CV

Suhwan Cho , Seoung Wug Oh , Sangyoun Lee , Joon-Young Lee This is my paper · ORCID
classification cs.CV
keywords videoinpaintingflow-guidedone-shotpixelpullingreferencegenerationStableDiffusionobjectremovalHQVIbenchmarkopticalflowpropagation
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

RGVI claims that video inpainting is best done by decoupling content propagation from content generation: first move known pixels between frames, then let a large generative model fill only what remains. The paper introduces one-shot pixel pulling, which chains completed optical flows across frames and warps flow maps rather than color values, so each target pixel is sampled exactly once and keeps sub-pixel precision without re-sampling artifacts. A Stable Diffusion reference generator fills the remaining holes in a single key frame chosen by connection count, and the generated pixels are propagated to every other frame through the same flow-warping operation. On the new HQVI benchmark, built by alpha-matte compositing at 2K resolution, RGVI reports the best object-removal scores among all compared methods, best perceptual quality with reference generation, and practical memory and runtime on 2K video.

What carries the argument

The load-bearing object is the one-shot pixel pulling protocol built on chained optical flows. Flows are computed with RAFT, erased inside the mask, completed with ProPainter's recurrent flow completion, and chained across frames by grid-warping the flow maps themselves: $f_{i\to j}=f_{i\to j-1}+w(f_{j-1\to j},\,f_{i\to j-1})$. The resulting correspondence map lets each target pixel pull its color once from the nearest source frames, avoiding the repeated sampling that blurs recurrent warping. A bi-directional forward/backward pass fills each target frame, and an L1 color-difference check between the two directions flags unreliable pixels for later stages. Reference generation then selects the key frame with the highest connection count $C_i=\sum_j\sum_p [w(\hat{M}_j,f_{i\to j})\odot \hat{M}_i]$, fills it with Stable Diffusion under a text prompt, propagates the generated content through the same flow-warping operation, and a lightweight convolutional network completes the residual holes.

What would settle it

Take a video in which a slowly moving target is occluded by a foreground object whose motion matches the background, so that flow completion plausibly aligns to the wrong texture; run RGVI with and without the positive mask and check whether the pulled pixels match the visible background at sub-pixel accuracy, and whether verification falsely accepts the wrong alignments. A cleaner quantitative version is to compare RGVI's propagation output on frames with known ground-truth optical flows, isolating how much error comes from flow completion rather than from the propagation or generation stages.

Watch

Extended reading notes

Core claim

The paper's central claim is that a decoupled flow-guided pipeline composed of one-shot pixel pulling plus generative reference filling is the most effective current recipe for video inpainting. In the object-removal setting on HQVI, RGVI without reference achieves the highest PSNR and SSIM among all compared methods at 240p and 480p, RGVI with reference achieves the best LPIPS and VFID, and at 1200x2160 the reference-augmented version wins on all four metrics. On the DAVIS and YouTube-VOS restoration benchmarks, RGVI without reference matches or exceeds the previous best flow-guided method. The paper attributes these results to warping optical flows instead of colors, which preserves sub-pixel accuracy while avoiding the error accumulation and texture blur of recurrent pixel warping, and to using a strong latent-diffusion generator only for the residual holes.

Load-bearing premise

The load-bearing premise is that the completed optical flows point to the correct source pixels: if a finished flow is silently wrong, one-shot pulling copies the wrong color, and the verification step only catches disagreement between the forward and backward passes, not misalignment on which both directions agree.

Editorial extensions

If this is right

  • On HQVI, RGVI outperforms all compared methods in the object-removal setting, with reference generation giving the best perceptual scores at 240p/480p and the best PSNR at 2K resolution.
  • One-shot pulling of colors via warped flow maps removes the re-sampling artifacts of recurrent pixel warping, so flow-guided inpainting can preserve fine texture over long propagation distances.
  • Decoupling generation from propagation makes video inpainting text-controllable: the same framework removes objects in the default mode and synthesizes new content from user prompts in generation mode.
  • The additional positive-mask trick reduces color-bleeding artifacts in occlusion scenarios by keeping the occluding object's original pixels intact.
  • HQVI provides a 2K, alpha-matte-composited benchmark with negative/positive masks and generation-required cases, which the paper presents as a more realistic evaluation setup than random masks.

Reading between the lines

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

  • My inference: because reference generation is a separate module, a stronger or more specialized image inpainting model could replace Stable Diffusion without retraining the propagation stage, so future image-inpainting advances should carry over to video inpainting almost directly.
  • My inference: the same flow-chaining and one-shot pulling scheme could benefit other video correspondence tasks such as frame interpolation, video object segmentation, or long-range matching, where sub-pixel accuracy and temporal consistency matter.
  • My inference: the verification step could be upgraded from color agreement to a learned confidence map, which might detect structural misalignment even when forward and backward flows agree on a wrong location.
  • My inference: HQVI's positive/negative mask annotations make it possible to benchmark occlusion handling separately from generation, which current datasets do not isolate.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. RGVI is a flow-guided video inpainting system that decouples content propagation from content generation. It completes masked optical flows with ProPainter's recurrent protocol, propagates known pixels by chaining flows and pulling colors once per target pixel, verifies forward/backward pulls with an L1 color threshold, generates a key-frame reference with Stable Diffusion under user text prompts, propagates the reference, and finishes with per-frame image inpainting. The paper introduces HQVI, a 1200x2160 benchmark built by alpha-matte compositing, and reports improved PSNR/SSIM/LPIPS/VFID over prior methods on HQVI plus DAVIS and YouTube-VOS, together with a 10-participant user study.

Significance. The one-shot pulling idea is the most distinctive contribution: warping flow maps rather than color values is a simple way to avoid re-sampling artifacts while retaining sub-pixel accuracy, and the controlled visual comparison in Figure 6 supports this benefit. The decoupled use of Stable Diffusion on a single key frame, with text-prompt controllability, is practical and scales to 2K resolution, where several baselines fail. The HQVI benchmark addresses a real gap in realistic, high-resolution object-removal evaluation. The main limitations are that the quantitative evidence rests heavily on a self-constructed benchmark without statistical testing, and that flow-completion accuracy and the verification threshold are not directly validated; these issues weaken but do not invalidate the central claim.

major comments (4)
  1. [Section 3.2, verification threshold] The verification rule accepts a forward/backward pull when the L1 distance of the three normalized color channels is below 1, i.e., an average per-channel disagreement of up to 1/3 is treated as reliable. The paper states that this threshold shows 'minimal observed variation' but provides no ablation or statistics on the fraction of invalidated pixels, so the threshold's robustness is not demonstrated. Since the verified one-shot pulling is the core propagation contribution and Table 3 compares it against recurrent warping as a whole, the benefit of one-shot pulling is not separated from the benefit of the threshold rule.
  2. [Section 3.2 and Table 3] The completed optical flows are never evaluated. The paper adopts ProPainter's recurrent flow completion and then relies on these flows for one-shot pulling, but the ablation in Table 3 holds the flows fixed and varies only the propagation protocol. Because HQVI contains ground-truth background frames, an experiment replacing completed flows with flows estimated from the ground-truth background would directly quantify how much of RGVI's performance depends on flow-completion accuracy. Without such an experiment, the claimed superiority of the propagation module is contingent on an external component whose error behavior is unmeasured.
  3. [Section 5.1, Tables 1-2 and Section 5.3, Table 5] No error bars, confidence intervals, significance tests, or per-sequence statistics are reported for any quantitative comparison, yet the abstract and Section 5.1 claim RGVI 'significantly outperforms' all existing methods. Some reported gaps are small (e.g., 240p PSNR 30.66 vs. 30.62 for RGVI vs. ProPainter, and LPIPS 0.0390 vs. 0.0388 for RGVI without reference vs. ProPainter), and the user study in Table 5 is based on average ranks from 10 participants without any statistical test. The evidence is therefore weaker than the language of the claims.
  4. [Section 4, HQVI dataset] The quantitative conclusions for object removal are drawn primarily from HQVI, a dataset constructed by the authors, but the paper does not state whether HQVI will be released and provides no details on dataset size, sequence lengths, or the number of occluding-object sequences. The abstract and GitHub link mention code and models only. If HQVI is not made available, the main benchmark result is not reproducible; if it is released, the paper should document the composition statistics and any selection criteria used when choosing Pexels background videos.
minor comments (5)
  1. [Equation (1)] The symbol f_i->j is used both for the completed adjacent-frame flows and for the chained long-range flows, and the warping operation w(A,B) is not defined precisely; please clarify which argument is the displacement field and which is the quantity being warped, and state the direction convention.
  2. [Section 5.1] The text writes 'DA VIS 2016' but the dataset is normally called DAVIS 2016; please standardize the spelling and define the 'DA VI' and 'YTVI' abbreviations at first use.
  3. [Section 5.3] For the recurrent warping baseline in Table 3, the sentence 'the initial frame is used as the key frame since connection counts cannot be calculated' should be expanded to explain why the proposed key-frame selection is inapplicable to recurrent warping.
  4. [Figure 2] The pipeline diagram is very small and the four components are hard to read; please enlarge the figure or restructure it so that the flow completion, internal propagation, reference generation, and per-frame completion stages are visually distinct.
  5. [Section 3.5] The positive mask is user-provided; the paper should state how this mask is obtained in practice and confirm that it was not used in the main comparisons of Tables 1 and 2, so that the reported improvements are not dependent on an additional input that baselines do not receive.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; RGVI's propagation and generation claims are evaluated against external benchmarks and external pretrained components.

full rationale

The paper's central claims—one-shot pixel pulling improves propagation quality and RGVI outperforms prior methods—rest on external, independent components and external benchmarks. Masked flows are produced by RAFT plus ProPainter's recurrent completion, both pretrained off-the-shelf tools; reference generation uses an external Stable Diffusion model. The HQVI benchmark is contributed by the authors, but its ground truth is alpha-matte compositing from VideoMatte240K and Pexels videos, independent of RGVI's training; evaluating a method on a self-created benchmark with external ground truth is not circular. The comparison metrics (PSNR/SSIM/LPIPS/VFID) and the DAVIS/YouTube-VOS external benchmarks provide independent support. The propagation ablation (Table 3) compares recurrent warping against one-shot pulling under the same completed flows, isolating the claimed mechanism rather than measuring a fitted constant. The verification threshold is stated as empirically set to 1, but the paper notes minimal observed variation across values; this is a robustness consideration, not a fitted prediction. The acknowledged reliance on optical-flow accuracy is a dependency and correctness risk, not an equivalence with the method's output: imperfect flows can degrade results, but the output is not defined as a function of that limitation. Self-citations (e.g., Oh et al. 2019, which shares two co-authors with this paper) appear only in related-work context and do not carry the paper's load-bearing argument. Accordingly, no circular step can be exhibited, and the derivation chain is self-contained for the purposes of this review.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The performance claim rests on the quality of external optical flow and flow completion, the heuristic that flow-map warping is more robust than color warping, the validity of L1 color disagreement as a reliability check, the generative power of Stable Diffusion, and the representativeness of the self-constructed HQVI benchmark.

free parameters (1)
  • Propagation verification threshold = 1.0 (normalized L1 color distance)
    Set empirically in Section 3.2 as the threshold for accepting forward-backward color agreement; the authors note minimal variation across values, but it is a hand-tuned numeric choice.
assumptions (6)
  • domain assumption RAFT optical flows are sufficiently accurate for adjacent frames.
    The entire propagation pipeline uses RAFT flows as the ground truth for motion; errors in these flows directly corrupt the propagated pixels.
  • domain assumption ProPainter's recurrent protocol produces reliable completed flows in masked regions.
    Section 3.2 erases and completes flows inside the mask; if completed flows point to wrong locations, the one-shot pulling fills the target with incorrect content.
  • domain assumption Warping flow maps is more robust to re-sampling artifacts than warping color values.
    Section 3.2 states optical flow is inherently smoother, making it more robust; this is a heuristic, not a proven theorem, and underlies the design choice.
  • domain assumption L1 color difference between forward and backward pulls is a valid indicator of propagation reliability.
    The verification protocol in Section 3.2 uses this difference to mark unreliable pixels; a different measure could give different results.
  • domain assumption Stable Diffusion can generate plausible reference content for missing areas.
    The reference generation stage depends on the generative power of Stable Diffusion; the Limitations section admits generated content may sometimes appear unnatural.
  • domain assumption HQVI alpha matte composition simulates realistic video editing scenarios.
    Section 4 constructs the benchmark by alpha compositing foregrounds over Pexels videos; if this does not capture real occlusions, lighting, or motion, the benchmark conclusions may not transfer.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Elevating Flow-Guided Video Inpainting with Reference Generation." pith.science (2026). https://pith.science/paper/PBAQMRNO

@misc{pith2026241208975,
  author       = {Pith},
  title        = {Pith review of: Elevating Flow-Guided Video Inpainting with Reference Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PBAQMRNO}},
  note         = {Machine review of arXiv:2412.08975}
}
read the original abstract

Video inpainting (VI) is a challenging task that requires effective propagation of observable content across frames while simultaneously generating new content not present in the original video. In this study, we propose a robust and practical VI framework that leverages a large generative model for reference generation in combination with an advanced pixel propagation algorithm. Powered by a strong generative model, our method not only significantly enhances frame-level quality for object removal but also synthesizes new content in the missing areas based on user-provided text prompts. For pixel propagation, we introduce a one-shot pixel pulling method that effectively avoids error accumulation from repeated sampling while maintaining sub-pixel precision. To evaluate various VI methods in realistic scenarios, we also propose a high-quality VI benchmark, HQVI, comprising carefully generated videos using alpha matte composition. On public benchmarks and the HQVI dataset, our method demonstrates significantly higher visual quality and metric scores compared to existing solutions. Furthermore, it can process high-resolution videos exceeding 2K resolution with ease, underscoring its superiority for real-world applications.

Figures

Figures reproduced from arXiv: 2412.08975 by the authors.

Figure 1
Figure 1. Qualitative comparison between RGVI and state-of-the-art methods. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Overall pipeline of RGVI. Flow-based methods typically decouple propagation and generation modules, often yielding clearer results compared to end-to-end methods. However, they are not without limi￾tations. For instance, FGVC’s per-pixel forward flow tracing lacks sub-pixel accuracy, leading to spatial misalignment. Conversely, E2FGVI and ProPainter achieve sub-pixel accu￾racy with a recurrent pixel warping approach… view at source ↗
Figure 3
Figure 3. RGVI outputs from the generation mode. other frames. Formally, the connection count Ci for frame i is computed as follows: Ci = X L j=1 nX p [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Example videos from the HQVI dataset. Nega [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: RGVI outputs on the video restoration scenarios. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Visual comparison between recurrent pixel warp [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Qualitative analysis of the use of additional masks. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 20 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Y.; Lee, K.-Y.; and Hsu, W

    Chang, Y.-L.; Liu, Z. Y.; Lee, K.-Y.; and Hsu, W. 2019 a . Free-form video inpainting with 3d gated convolution and temporal patchgan. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 9066--9075

    work page 2019

  4. [4]

    Learnable Gated Temporal Shift Module for Deep Video Inpainting

    Chang, Y.-L.; Liu, Z. Y.; Lee, K.-Y.; and Hsu, W. 2019 b . Learnable gated temporal shift module for deep video inpainting. arXiv preprint arXiv:1907.01131

    work page Pith review arXiv 2019

  5. [5]

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

    arXiv 2020

  6. [6]

    Gao, C.; Saraf, A.; Huang, J.-B.; and Kopf, J. 2020. Flow-edge guided video completion. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XII 16, 713--729. Springer

    work page 2020

  7. [7]

    B.; Ahuja, N.; and Kopf, J

    Huang, J.-B.; Kang, S. B.; Ahuja, N.; and Kopf, J. 2016. Temporally coherent completion of dynamic video. ACM Transactions on Graphics (ToG), 35(6): 1--11

    work page 2016

  8. [8]

    W.; and Kim, S

    Kang, J.; Oh, S. W.; and Kim, S. J. 2022. Error compensation framework for flow-guided video inpainting. In European Conference on Computer Vision, 375--390. Springer

    work page 2022

Show all 39 references
  1. [9]

    Kim, D.; Woo, S.; Lee, J.-Y.; and Kweon, I. S. 2019. Deep video inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 5792--5801

    2019

  2. [10]

    P.; and Ba, J

    Kingma, D. P.; and Ba, J. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980

    2014 arXiv

  3. [11]

    Krizhevsky, A.; Sutskever, I.; and Hinton, G. E. 2012. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25

    2012

  4. [12]

    Lao, D.; Zhu, P.; Wonka, P.; and Sundaramoorthi, G. 2021. Flow-guided video inpainting with scene templates. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 14599--14608

    2021

  5. [13]

    W.; Won, D.; and Kim, S

    Lee, S.; Oh, S. W.; Won, D.; and Kim, S. J. 2019. Copy-and-paste networks for deep video inpainting. In Proceedings of the IEEE/CVF international conference on computer vision, 4413--4421

    2019

  6. [14]

    Li, A.; Zhao, S.; Ma, X.; Gong, M.; Qi, J.; Zhang, R.; Tao, D.; and Kotagiri, R. 2020. Short-term and long-term context aggregation network for video inpainting. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part IV 16,...

    2020

  7. [15]

    Li, Z.; Lu, C.-Z.; Qin, J.; Guo, C.-L.; and Cheng, M.-M. 2022. Towards an end-to-end framework for flow-guided video inpainting. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 17562--17571

    2022

  8. [16]

    L.; Seitz, S

    Lin, S.; Ryabtsev, A.; Sengupta, S.; Curless, B. L.; Seitz, S. M.; and Kemelmacher-Shlizerman, I. 2021. Real-time high-resolution background matting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8762--8771

    2021

  9. [17]

    A.; Shih, K

    Liu, G.; Reda, F. A.; Shih, K. J.; Wang, T.-C.; Tao, A.; and Catanzaro, B. 2018. Image inpainting for irregular holes using partial convolutions. In Proceedings of the European conference on computer vision (ECCV), 85--100

    2018

  10. [18]

    Liu, R.; Deng, H.; Huang, Y.; Shi, X.; Lu, L.; Sun, W.; Wang, X.; Dai, J.; and Li, H. 2021. Fuseformer: Fusing fine-grained information in transformers for video inpainting. In Proceedings of the IEEE/CVF international conference on computer vision, 14040--14049

    2021

  11. [19]

    W.; Lee, S.; Lee, J.-Y.; and Kim, S

    Oh, S. W.; Lee, S.; Lee, J.-Y.; and Kim, S. J. 2019. Onion-peel networks for deep video completion. In Proceedings of the IEEE/CVF international conference on computer vision, 4403--4412

    2019

  12. [20]

    Ouyang, H.; Wang, T.; and Chen, Q. 2021. Internal video inpainting by implicit long-range propagation. In Proceedings of the IEEE/CVF international conference on computer vision, 14579--14588

    2021

  13. [21]

    Perazzi, F.; Pont-Tuset, J.; McWilliams, B.; Van Gool, L.; Gross, M.; and Sorkine-Hornung, A. 2016. A benchmark dataset and evaluation methodology for video object segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 724--732

    2016

  14. [22]

    Ren, J.; Zheng, Q.; Zhao, Y.; Xu, X.; and Li, C. 2022. Dlformer: Discrete latent transformer for video inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3511--3520

    2022

  15. [23]

    Rombach, R.; Blattmann, A.; Lorenz, D.; Esser, P.; and Ommer, B. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 10684--10695

    2022

  16. [24]

    Teed, Z.; and Deng, J. 2020. Raft: Recurrent all-pairs field transforms for optical flow. In European conference on computer vision, 402--419. Springer

    2020

  17. [25]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. Advances in neural information processing systems, 30

    2017

  18. [26]

    Wang, C.; Huang, H.; Han, X.; and Wang, J. 2019. Video inpainting by jointly learning temporal structure and spatial details. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33, 5232--5239

    2019

  19. [27]

    Wang, T.-C.; Liu, M.-Y.; Zhu, J.-Y.; Liu, G.; Tao, A.; Kautz, J.; and Catanzaro, B. 2018. Video-to-video synthesis. arXiv preprint arXiv:1808.06601

    2018 arXiv

  20. [28]

    C.; Sheikh, H

    Wang, Z.; Bovik, A. C.; Sheikh, H. R.; and Simoncelli, E. P. 2004. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4): 600--612

    2004

  21. [29]

    Wu, Z.; Xuan, H.; Sun, C.; Guan, W.; Zhang, K.; and Yan, Y. 2023. Semi-Supervised Video Inpainting with Cycle Consistency Constraints. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 22586--22595

    2023

  22. [30]

    Xu, N.; Yang, L.; Fan, Y.; Yue, D.; Liang, Y.; Yang, J.; and Huang, T. 2018. Youtube-vos: A large-scale video object segmentation benchmark. arXiv preprint arXiv:1809.03327

    2018 arXiv

  23. [31]

    Xu, R.; Li, X.; Zhou, B.; and Loy, C. C. 2019. Deep flow-guided video inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3723--3732

    2019

  24. [32]

    Yu, J.; Lin, Z.; Yang, J.; Shen, X.; Lu, X.; and Huang, T. S. 2018. Generative image inpainting with contextual attention. In Proceedings of the IEEE conference on computer vision and pattern recognition, 5505--5514

    2018

  25. [33]

    Yu, J.; Lin, Z.; Yang, J.; Shen, X.; Lu, X.; and Huang, T. S. 2019. Free-form image inpainting with gated convolution. In Proceedings of the IEEE/CVF international conference on computer vision, 4471--4480

    2019

  26. [34]

    Zeng, Y.; Fu, J.; and Chao, H. 2020. Learning joint spatial-temporal transformations for video inpainting. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XVI 16, 528--543. Springer

    2020

  27. [35]

    Zhang, H.; Mai, L.; Xu, N.; Wang, Z.; Collomosse, J.; and Jin, H. 2019. An internal learning approach to video inpainting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 2720--2729

    2019

  28. [36]

    Zhang, K.; Fu, J.; and Liu, D. 2022. Flow-guided transformer for video inpainting. In European Conference on Computer Vision, 74--90. Springer

    2022

  29. [37]

    A.; Shechtman, E.; and Wang, O

    Zhang, R.; Isola, P.; Efros, A. A.; Shechtman, E.; and Wang, O. 2018. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recognition, 586--595

    2018

  30. [38]

    C.; and Loy, C

    Zhou, S.; Li, C.; Chan, K. C.; and Loy, C. C. 2023. ProPainter: Improving Propagation and Transformer for Video Inpainting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 10477--10486

    2023

  31. [39]

    Zou, X.; Yang, L.; Liu, D.; and Lee, Y. J. 2021. Progressive temporal feature alignment network for video inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 16448--16457

    2021

Pith tools

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