The DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.
Video Demoiréing with Relation-Based Temporal Consistency
Abstract
Moiré patterns, appearing as color distortions, severely degrade image and video qualities when filming a screen with digital cameras. Considering the increasing demands for capturing videos, we study how to remove such undesirable moiré patterns in videos, namely video demoiréing. To this end, we introduce the first hand-held video demoiréing dataset with a dedicated data collection pipeline to ensure spatial and temporal alignments of captured data. Further, a baseline video demoiréing model with implicit feature space alignment and selective feature aggregation is developed to leverage complementary information from nearby frames to improve frame-level video demoiréing. More importantly, we propose a relation-based temporal consistency loss to encourage the model to learn temporal consistency priors directly from ground-truth reference videos, which facilitates producing temporally consistent predictions and effectively maintains frame-level qualities. Extensive experiments manifest the superiority of our model. Code is available at ht tps:// daipengwa. github.io/VDmoire_ProjectPage/.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
Most existing demoiréing methods struggle to remove large-scale, globally distributed moiré patterns in both images and videos.
“Second, most existing demoir´ eing methods [7, 1, 13, 5, 16, 17, 18] struggle to remove large-scale, globally distributed moir´ e patterns in both images and videos.”
From MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing (2026) · cites this paper as [16]
Applying image demoiréing models directly to video sequences often introduces flickering artifacts due to the absence of temporal consistency mechanisms.
“When applied directly to video sequences, they often introduce flickering artifacts due to the absence of temporal consistency mechanisms [16, 17, 18, 19, 26].”
From MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing (2026) · cites this paper as [16]
A transformer-based video demoiréing approach with sparse attention has heavier computational demand than CNN-based models.
“Although a transformer-based approach [25] has been explored in the video demoir´ eing domain to exploit global information [19], its sparse attention mechanism still results in heavier computational demand compared to CNN-based models [16, 17, 18].”
From MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing (2026) · cites this paper as [16]
Moiré patterns introduce significant color distortion.
“Moir´ e patterns not only obscure image content with their characteristic grid-like structures but also introduce significant color distortion [26, 17, 16].”
From MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing (2026) · cites this paper as [16]
Citations
Cites
U-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404
Deep High-Resolution Representation Learning for Visual Recognition
Wang, J., Sun, K., Cheng, T., Jiang, B., Deng, C., Zhao, Y., Liu, D., Mu, Y., Tan, M., Wang, X., et al. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2020.2983686 · Cited by 4,894
Wang, T.-C., Liu, M.-Y., Zhu, J.-Y., Liu, G., Tao, A., Kautz, J., & Catanzaro, B. (2018) · arXiv · doi:10.48550/arxiv.1808.06601 · Cited by 594
Flagged first, then by citations
Research works printed but not linked: 51
- 10.1109/CVPR.2017.33 · DOI
- sidorov, 0, Suppression of moiré patterns via spectral analysis, VCIP
- 10.1145/3343031.3350864 · DOI
- 10.1109/ICCV48922.2021.01431 · DOI
- nishioka, 2000, Endoscope system provided with low-pass filter for moire removal
- 10.1201/b12081 · DOI
- 10.1109/CVPR.2018.00693 · DOI
- 10.1109/ICCV.2011.6126544 · DOI · Paper page
- pharr, 2016, Physically Based Rendering From Theory to Implementation
- 10.1109/CVPRW53098.2021.00058 · DOI
- loshchilov, 2016, Sgdr: Stochastic gradient descent with warm restarts, ArXiv Preprint
- 10.1109/CVPRW.2019.00251 · DOI
- 10.1145/2816795.2818107 · DOI
- 10.1109/ICPR.2000.903475 · DOI
- lee, 0, Deeply-supervised nets, Artificial Intelligence and Statistics
- lei, 0, Blind video temporal consistency via deep video prior, NeurIPS
- 10.1109/CVPR.2019.00387 · DOI
- 10.1109/CVPR.2018.00341 · DOI · Paper page
- liu, 0, Moiré pattern removal from texture images via low-rank and sparse matrix decomposition, VCIP
- liu, 2020, Self-adaptively learning to demoiré from focused and defocused image pairs, ArXiv Preprint
- liu, 0, Wavelet-based dual-branch network for image demoiréing, ECCV
- 10.1109/CVPR42600.2020.00583 · DOI
- 10.1109/CVPR46437.2021.00493 · DOI
- 10.1109/CVPR42600.2020.00369 · DOI
- 10.1109/CVPR.2018.00068 · DOI · Paper page
- 10.1109/ICCV.2019.00281 · DOI
- 10.1109/ICCV.2017.89 · DOI · Paper page
- 10.3758/BF03193280 · DOI
- sun, 2018, Moiré photo restoration using multiresolution convolutional neural networks, TIP
- 10.1109/ICCV.2015.316 · DOI · Paper page
- 10.1109/CVPR.2019.01143 · DOI
- 10.1109/ICCV.2019.00251 · DOI · Paper page
- he, 0, Fhde 2 net: Full high definition demoireing network, ECCV
- 10.1109/CVPR.2017.179 · DOI
- johnson, 0, Perceptual losses for real-time style transfer and super-resolution, ECCV
- lai, 0, Learning blind video temporal consistency, ECCV
- lamdouar, 0, Betrayed by motion: Camouflaged object discovery via motion segmentation, ACCV
- 10.1109/CVPR.2017.502 · DOI · Paper page
- 10.1109/CVPR.2017.304 · DOI
- chan, 2021, Basicvsr++: Improving video super-resolution with enhanced propagation and alignment, ArXiv Preprint
- 10.1109/CVPR46437.2021.00491 · DOI
- 10.1088/1748-3190/aaffb5 · DOI
- 10.1109/ICCV.2019.00916 · DOI
- 10.1109/CVPRW.2019.00247 · DOI · Paper page
- 10.1016/j.image.2018.06.009 · DOI
- 10.1145/1873951.1874249 · DOI
- 10.1609/aaai.v34i07.6905 · DOI
- teed, 0, Raft: Recurrent all-pairs field transforms for optical flow, ECCV
- 10.1109/CVPR42600.2020.00143 · DOI
- tsai, 0, Multiframe image restoration and registration, Advance Computer Visual and Image Processing
- 10.1109/CVPR42600.2020.00342 · DOI
Cited by
Sorted by citations
Std-net: spatio-temporal decomposition network for video demoiréing with sparse transformers
Niu, Y., Xu, R., Lin, Z., & Liu, W. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3386604 · Cited by 12
MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing
Sung, J., Roh, C., Eom, C., & Oh, J. (2026) · Neurocomputing · doi:10.1016/j.neucom.2025.132477
Metadata
- Authors
- Dai, P., Yu, X., Ma, L., Zhang, B., Li, J., Li, W., Shen, J., & Qi, X.
- Affiliations
- University of Hong Kong (HK)
- TCL (China) (CN)
- Sun Yat-sen University (CN)
- Chinese University of Hong Kong (HK)
- Venue
- 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2022-06
- Pages
- 17601-17610
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W4312354721
- Cited by
- 26
- References
- 3
- Citations by year
- 2026: 3 · 2025: 8 · 2024: 9 · 2023: 5 · 2022: 1
- FWCI
- 2.08
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image Processing Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Enhancement Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Keywords
- feature space alignment, color distortion
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
