CoChat
Sign Up
Cited by 604Advanced Image Processing TechniquesGenerative Adversarial Networks and Image SynthesisImage Enhancement Techniques

The DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.

BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation and Alignment

Open access
Chan, Kelvin C.K., Zhou, Shangchen, Xu, Xiangyu, Loy, Chen Change·2022·2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
DOI 10.1109/cvpr52688.2022.00588
Open this paper in CoChat — free accountOpen PDF

Abstract

A recurrent structure is a popular framework choice for the task of video super-resolution. The state-of-the-art method BasicVSR adopts bidirectional propagation with feature alignment to effectively exploit information from the entire input video. In this study, we redesign BasicVsr by proposing second-order grid propagation and flow-guided deformable alignment. We show that by empowering the re-current framework with enhanced propagation and align-ment, one can exploit spatiotemporal information across misaligned video frames more effectively. The new components lead to an improved performance under a simi-lar computational constraint. In particular, our model Ba-sicVSR++ surpasses BasicVSR by a significant 0.82 dB in PSNR with similar number of parameters. BasicVSR++ is generalizable to other video restoration tasks, and obtains three champions and one first runner-up in NTIRE 2021 video restoration challenge.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/cvpr52688.2022.00588

Cites

3 of 3 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Recent video demoiréing methods using PCD alignment and temporal consistency loss involve significant parameter overhead and are prone to training instability.

    “However, these approaches typically involve significant parameter overhead and are prone to training instability [27, 28].”

    From MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing (2026) · cites this paper as [27]

    Section 3.2 says deformable alignment can be difficult to train and that instability can cause training to collapse. The Introduction and experimental comparison also describe sliding-window VSR models as less compact, with EDVR having more parameters than BasicVSR++. But the paper studies video super-resolution, not video demoiréing, and it does not attribute parameter overhead or training instability to temporal-consistency loss. (LLM, conf=0.88)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 37

  1. 10.1109/ICCV.2019.00257 · DOI
  2. 10.1109/CVPR42600.2020.00343 · DOI
  3. 10.1109/ICCVW54120.2021.00217 · DOI
  4. 10.1109/CVPRW.2019.00247 · DOI · Paper page
  5. yi, 2019, Progressive fusion video super-resolution net-work via exploiting non-local spatio-temporal correlations, ICCV
  6. 10.1109/CVPRW53098.2021.00075 · DOI
  7. 10.1007/s11263-018-01144-2 · DOI
  8. 10.1109/TIP.2020.2999209 · DOI
  9. huang, 2015, Bidirectional recurrent convolutional networks for multi-frame super-resolution, NIPS
  10. 10.1109/ICCVW.2019.00113 · DOI
  11. 10.1109/TPAMI.2017.2701380 · DOI
  12. isobe, 2020, Video super-resolution with recurrent structure-detail network, ECCV
  13. 10.1109/CVPR42600.2020.00803 · DOI
  14. isobe, 2020, Revisiting temporal modeling for video super-resolution, BMVC
  15. jo, 2018, Deep video super-resolution network using dy-namic upsampling filters without explicit motion compensation, CVPR
  16. rosemary ke, 2018, Sparse attentive backtracking: Temporal creditassignment through reminding, NIPS
  17. li, 2020, MuCAN: Multi-correspondence aggregation network for video super-resolution, ECCV
  18. lin, 1996, Learning long-term dependencies in NARX recurrent neural networks, IEEE Transactions on Neural Networks
  19. liu, 2014, On bayesian adaptive video super resolution, TPAMI
  20. tian, 2018, TDAN: Temporally deformable alignment network for video super-resolution, CVPR
  21. 10.1609/aaai.v35i2.16181 · DOI · Paper page
  22. 10.1109/ICCV.2017.479 · DOI
  23. 10.1109/CVPR46437.2021.00491 · DOI
  24. 10.1109/ACCESS.2019.2958030 · DOI
  25. 10.1109/CVPR52688.2022.00587 · DOI
  26. 10.1109/ICCVW.2019.00431 · DOI
  27. fourure, 2017, Residual conv-deconv grid network for semantic segmentation, BMVC
  28. cao, 2021, Video super-resolution transformer, ArXiv Preprint
  29. 10.1109/CVPR.2019.00402 · DOI
  30. 10.1109/CVPR.2017.304 · DOI
  31. 2022, Contributors MMEditing, MMEditing OpenMMLab Image and Video Editing Toolbox
  32. nah, 2019, Re-current neural networks with intra-frame iterations for video deblurring, CVPR
  33. 10.1109/CVPRW.2019.00251 · DOI
  34. 10.1109/CVPR.2018.00693 · DOI
  35. 10.1109/CVPR42600.2020.00548 · DOI
  36. 10.29007/21r5 · DOI
  37. soltani, 2016, Higher order recurrent neu-ral networks, ArXiv Preprint

Cited by

Sorted by citations

Metadata

Authors
Chan, K. C. K., Zhou, S., Xu, X., & Loy, C. C.
Affiliations
Nanyang Technological University (SG)
ORCID
Venue
2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Publisher
IEEE
Published
2022-06
Pages
5962-5971
ISSN
1063-6919 (ISSN-L)
DOI
10.1109/cvpr52688.2022.00588
OpenAlex
W3157035436
Cited by
604
References
3
Citations by year
2026: 104 · 2025: 147 · 2024: 162 · 2023: 135 · 2022: 51
FWCI
44.87
Type
conference paper
Language
en
Open access
Licence
Locations
  • IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
  • arXiv (Cornell University) (submittedVersion, cc-by)
  • arXiv (Cornell University) (cc-by)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image Processing Techniques
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Generative Adversarial Networks and Image Synthesis
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Enhancement Techniques
Keywords
video super-resolution, recurrent neural networks, bidirectional propagation, feature alignment, spatio-temporal data, video restoration

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review