Papers › VRT: A Video Restoration Transformer

VRT: A Video Restoration Transformer

28 Jan 2022arXiv:2201.12288archive 2025-07-28

Jingyun Liang, JieZhang Cao, Yuchen Fan, Kai Zhang, Rakesh Ranjan, Yawei Li, Radu Timofte, Luc van Gool

Video restoration (e.g., video super-resolution) aims to restore high-quality frames from low-quality frames. Different from single image restoration, video restoration generally requires to utilize temporal information from multiple adjacent but usually misaligned video frames. Existing deep methods generally tackle with this by exploiting a sliding window strategy or a recurrent architecture, which either is restricted by frame-by-frame restoration or lacks long-range modelling ability. In this paper, we propose a Video Restoration Transformer (VRT) with parallel frame prediction and long-range temporal dependency modelling abilities. More specifically, VRT is composed of multiple scales, each of which consists of two kinds of modules: temporal mutual self attention (TMSA) and parallel warping. TMSA divides the video into small clips, on which mutual attention is applied for joint motion estimation, feature alignment and feature fusion, while self attention is used for feature extraction. To enable cross-clip interactions, the video sequence is shifted for every other layer. Besides, parallel warping is used to further fuse information from neighboring frames by parallel feature warping. Experimental results on five tasks, including video super-resolution, video deblurring, video denoising, video frame interpolation and space-time video super-resolution, demonstrate that VRT outperforms the state-of-the-art methods by large margins (up to 2.16dB) on fourteen benchmark datasets.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 5 code samples: 3 from 1 repository linked to this paper, 2 identical to code harvested from another paper's repository. Across all 5: 4 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 1 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 3 where the contract check could not run; failures of Syntology's instrument, not of the code. Every run on this page is a failure of Syntology's instrument, not of the code.

By repository: official repository: 3 samples from 1 repository, 3 ran (0 with no instrument failure; 3 where Syntology's instrument failed); 2 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

jingyunliang/vrt officialmentioned in papermentioned on GitHubpytorchSyntology: 3 samples, 3 ran (3 instrument-failed) · 0 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

5 samples harvested: 4 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 1 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Every run on this page is a failure of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · our draft was wrong
3ran; the contract check could not run
1unverified

Licence: 5 of the 5 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from jingyunliang/vrt. 2 samples are listed last, in a group without a repository. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

jingyunliang/vrt official repository 3 samples, 3 ran (3 instrument-failed) · 0 unverified
drop_path in 325 papers
census: 27 implementations from 61 papers, 5 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 4, kind float, dtype float32
models/network_vrt.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant licence not identified · pointer only · 9ea4d5b8b4d5ef2f · report fixture: signature x:Tensor/4/float, drop_prob:float/0/float, training:bool/0/bool · returned Tensor · level recorded 2026-09-22
flow_warp in 48 papers models/network_vrt.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant licence not identified · pointer only · c4d261acbb02d0e1 · report fixture: signature x:Tensor/4/float, flow:Tensor/4/float, interp_mode:str/0/str, padding_mode:str/0/str, align_corners:bool/0/bool, use_pad_mask:bool/0/bool · returned Tensor · level recorded 2026-09-22
trunc_normal_ in 154 papers
census: 37 implementations from 54 papers, 4 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
models/network_vrt.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant licence not identified · pointer only · 8347c57b16dab539 · report fixture: signature tensor:Tensor/2/float, mean:float/0/float, std:float/0/float, a:float/0/float, b:float/0/float · returned Tensor · level recorded 2026-09-22
Repository not located 2 samples, 1 ran (1 instrument-failed) · 1 unverified
2 samples are identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
test_video repository not located ran · our draft was wrong licence of this copy not recorded · a1cad82b9beff0c7 · report fixture: signature lq:Tensor/5/float, model:callable/0/?, args:Args/0/? · returned Tensor · level recorded 2026-09-22
test_clip repository not located unverified licence of this copy not recorded · 705758eeba9070e3 · report

Tasks

DeblurringDenoisingImage RestorationMotion EstimationSpace-time Video Super-resolutionSuper-ResolutionVideo DeblurringVideo DenoisingVideo Frame InterpolationVideo RestorationVideo Super-Resolution

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Deblurring BASED VRT (GoPro) PSNR 31.42945 #1 of 4 Archive leaderboard report
Deblurring BASED VRT (GoPro) VMAF 66.72253 #1 of 4 Archive leaderboard report
Deblurring BASED VRT (REDS) ERQAv2.0 0.74874 #3 of 4 Archive leaderboard report
Deblurring BASED VRT (REDS) LPIPS 0.08248 #3 of 4 Archive leaderboard report
Deblurring BASED VRT (REDS) PSNR 30.97878 #3 of 4 Archive leaderboard report
Deblurring BASED VRT (REDS) SSIM 0.94503 #3 of 4 Archive leaderboard report
Deblurring BASED VRT (REDS) VMAF 66.81782 #3 of 4 Archive leaderboard report
Deblurring DVD VRT PSNR 34.27 #4 of 7 Archive leaderboard report
Deblurring GoPro VRT PSNR 34.81 #6 of 56 Archive leaderboard report
Deblurring GoPro VRT SSIM 0.9724 #6 of 56 Archive leaderboard report
Deblurring MSU BASED VRT (GoPro) ERQAv2.0 0.74874 #2 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (GoPro) LPIPS 0.08165 #2 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (GoPro) PSNR 31.42945 #2 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (GoPro) SSIM 0.94503 #2 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (GoPro) Subjective 2.3854 #2 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (GoPro) VMAF 66.72253 #2 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (REDS) ERQAv2.0 0.75056 #3 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (REDS) LPIPS 0.08248 #3 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (REDS) PSNR 30.97878 #3 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (REDS) SSIM 0.94601 #3 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (REDS) Subjective 1.5660 #3 of 11 Archive leaderboard report
Deblurring MSU BASED VRT (REDS) VMAF 66.81782 #3 of 11 Archive leaderboard report
Deblurring REDS VRT Average PSNR 36.79 #1 of 3 Archive leaderboard report
Space-time Video Super-resolution Vimeo90K-Fast VRT PSNR 36.98 #1 of 4 Archive leaderboard report
Space-time Video Super-resolution Vimeo90K-Fast VRT SSIM 0.9439 #1 of 4 Archive leaderboard report
Space-time Video Super-resolution Vimeo90K-Medium VRT PSNR 36.01 #1 of 4 Archive leaderboard report
Space-time Video Super-resolution Vimeo90K-Medium VRT SSIM 0.9434 #1 of 4 Archive leaderboard report
Video Denoising DAVIS sigma10 VRT PSNR 40.82 #2 of 6 Archive leaderboard report
Video Denoising DAVIS sigma20 VRT PSNR 38.15 #2 of 7 Archive leaderboard report
Video Denoising DAVIS sigma30 VRT PSNR 36.52 #3 of 7 Archive leaderboard report
Video Denoising DAVIS sigma40 VRT PSNR 35.32 #3 of 8 Archive leaderboard report
Video Denoising DAVIS sigma50 VRT PSNR 34.36 #3 of 8 Archive leaderboard report
Video Denoising Set8 sigma10 VRT PSNR 37.88 #2 of 6 Archive leaderboard report
Video Denoising Set8 sigma20 VRT PSNR 35.02 #2 of 7 Archive leaderboard report
Video Denoising Set8 sigma30 VRT PSNR 33.35 #2 of 7 Archive leaderboard report
Video Denoising Set8 sigma40 VRT PSNR 32.15 #3 of 8 Archive leaderboard report
Video Denoising Set8 sigma50 VRT PSNR 31.22 #3 of 9 Archive leaderboard report
Video Frame Interpolation Vid4 - 4x upscaling VRT PSNR 27.46 #1 of 5 Archive leaderboard report
Video Frame Interpolation Vid4 - 4x upscaling VRT Parameters 4450000 #1 of 5 Archive leaderboard report
Video Frame Interpolation Vid4 - 4x upscaling VRT SSIM 0.8392 #1 of 5 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x264 BSQ-rate over ERQA 1.578 #17 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x264 BSQ-rate over LPIPS 1.259 #17 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x264 BSQ-rate over MS-SSIM 0.662 #17 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x264 BSQ-rate over PSNR 1.09 #17 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x264 BSQ-rate over Subjective Score 1.245 #17 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x264 BSQ-rate over VMAF 0.7 #17 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x265 BSQ-rate over ERQA 8.92 #28 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x265 BSQ-rate over LPIPS 11.329 #28 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x265 BSQ-rate over MS-SSIM 1.257 #28 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x265 BSQ-rate over PSNR 6.634 #28 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x265 BSQ-rate over Subjective Score 2.023 #28 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + x265 BSQ-rate over VMAF 1.217 #28 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + vvenc BSQ-rate over ERQA 18.333 #29 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + vvenc BSQ-rate over LPIPS 11.496 #29 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + vvenc BSQ-rate over MS-SSIM 0.836 #29 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + vvenc BSQ-rate over PSNR 5.777 #29 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + vvenc BSQ-rate over Subjective Score 2.235 #29 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + vvenc BSQ-rate over VMAF 0.652 #29 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + uavs3e BSQ-rate over ERQA 6.619 #33 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + uavs3e BSQ-rate over LPIPS 4.003 #33 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + uavs3e BSQ-rate over MS-SSIM 1.982 #33 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + uavs3e BSQ-rate over PSNR 5.862 #33 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + uavs3e BSQ-rate over Subjective Score 2.511 #33 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + uavs3e BSQ-rate over VMAF 1.425 #33 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + aomenc BSQ-rate over ERQA 12.289 #34 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + aomenc BSQ-rate over LPIPS 4.429 #34 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + aomenc BSQ-rate over MS-SSIM 2.797 #34 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + aomenc BSQ-rate over PSNR 10.075 #34 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + aomenc BSQ-rate over Subjective Score 2.631 #34 of 85 Archive leaderboard report
Video Super-Resolution MSU Super-Resolution for Video Compression VRT + aomenc BSQ-rate over VMAF 1.733 #34 of 85 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT 1 - LPIPS 0.929 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT ERQAv1.0 0.758 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT FPS 2.778 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT PSNR 31.669 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT QRCRv1.0 0.722 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT SSIM 0.902 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Super Resolution Benchmark: Detail Restoration VRT Subjective score 7.628 #1 of 32 Archive leaderboard report
Video Super-Resolution MSU Video Upscalers: Quality Enhancement VRT-Reds-L LPIPS 0.343 #29 of 48 Archive leaderboard report
Video Super-Resolution MSU Video Upscalers: Quality Enhancement VRT-Reds-L PSNR 31.01 #29 of 48 Archive leaderboard report
Video Super-Resolution MSU Video Upscalers: Quality Enhancement VRT-Reds-L SSIM 0.869 #29 of 48 Archive leaderboard report
Video Super-Resolution UDM10 - 4x upscaling VRT PSNR 41.05 #1 of 7 Archive leaderboard report
Video Super-Resolution UDM10 - 4x upscaling VRT SSIM 0.9737 #1 of 7 Archive leaderboard report
Video Super-Resolution Vid4 - 4x upscaling VRT PSNR 27.93 #8 of 27 Archive leaderboard report
Video Super-Resolution Vid4 - 4x upscaling VRT SSIM 0.8425 #8 of 27 Archive leaderboard report
Video Super-Resolution Vid4 - 4x upscaling - BD degradation VRT PSNR 29.42 #2 of 18 Archive leaderboard report
Video Super-Resolution Vid4 - 4x upscaling - BD degradation VRT SSIM 0.8795 #2 of 18 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections