Papers › ProPainter: Improving Propagation and Transformer for Video Inpainting
ProPainter: Improving Propagation and Transformer for Video Inpainting
Shangchen Zhou, Chongyi Li, Kelvin C. K. Chan, Chen Change Loy
Flow-based propagation and spatiotemporal Transformer are two mainstream mechanisms in video inpainting (VI). Despite the effectiveness of these components, they still suffer from some limitations that affect their performance. Previous propagation-based approaches are performed separately either in the image or feature domain. Global image propagation isolated from learning may cause spatial misalignment due to inaccurate optical flow. Moreover, memory or computational constraints limit the temporal range of feature propagation and video Transformer, preventing exploration of correspondence information from distant frames. To address these issues, we propose an improved framework, called ProPainter, which involves enhanced ProPagation and an efficient Transformer. Specifically, we introduce dual-domain propagation that combines the advantages of image and feature warping, exploiting global correspondences reliably. We also propose a mask-guided sparse video Transformer, which achieves high efficiency by discarding unnecessary and redundant tokens. With these components, ProPainter outperforms prior arts by a large margin of 1.46 dB in PSNR while maintaining appealing efficiency.
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):
get_harvested_code_for_paper(arxiv_id="2309.03897")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2309.03897", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2309.03897")
Code
Syntology 34 code samples harvested from 2 repositories linked to this paper: 20 ran (of which 10 constructed an object rather than computing a result; 12 with no instrument failure: 0 honoured, 0 violated, 12 with no contract checked; 8 where Syntology's instrument failed) · 14 unverified. Where Syntology's instrument failed: 6 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.
By repository: official repository: 16 samples from 1 repository, 10 ran (6 with no instrument failure; 4 where Syntology's instrument failed); community (archive-listed): 18 samples from 1 repository, 10 ran (6 with no instrument failure; 4 where Syntology's instrument failed). 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 between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.
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
34 samples harvested: 20 ran (of which 10 constructed an object rather than computing a result; 12 with no instrument failure: 0 honoured, 0 violated, 12 with no contract checked; 8 where Syntology's instrument failed) · 14 unverified. “Where Syntology's instrument failed” counts failures 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.
Licence: 16 of the 34 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 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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”. For 1 sample that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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.
4cdb3a50bc7f846f · report
fixture: signature x:Tensor/4/float, cond_feat:Tensor/4/float, flow:Tensor/4/float · returned Tensor · level recorded 2026-09-24 53a5eb8ddb64103e · report
fixture: returned Encoder · level recorded 2026-09-22 83d5cf110edb5af1 · report
fixture: returned FusionFeedForward · level recorded 2026-09-22 cbeb924f2247e025 · report
fixture: returned ModulatedDeformConv2d · level recorded 2026-09-22 9ff5a33dc359eb58 · report
fixture: returned SoftSplit · level recorded 2026-09-22 c4696b981f65032e · report
fixture: returned SparseWindowAttention · level recorded 2026-09-22 e34f0d309218c4cb · report
fixture: signature flow_fw:Tensor/4/float, flow_bw:Tensor/4/float, alpha1:float/0/float, alpha2:float/0/float · returned Tensor · level recorded 2026-09-22 0724aee9bf22b8ca · report
fixture: signature x:Tensor/4/float, flow:Tensor/4/float, interpolation:str/0/str, padding_mode:str/0/str, align_corners:bool/0/bool · returned Tensor · level recorded 2026-09-22 99919ce37333a615 · report
fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22 census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket model/propainter.py ran; the contract check could not run (RAISES) licence not identified · pointer only ·
c384ddd8232a9399 · report
fixture: signature x:Tensor/5/float, window_size:tuple/1/int, n_head:int/0/int · returned Tensor · level recorded 2026-09-22 a7a87d9746b62c93 · report
7b9c10bbcaf37395 · report
ae3ca4ef5eac8aec · report
0355e712ec5750ab · report
fa4fe301319a48a2 · report
ae01e24194ffa4ed · report
83365c19252ece6f · report
fixture: returned FusionFeedForward · level recorded 2026-09-22 2fd5dd72e6bde9a9 · report
fixture: returned InterpolationBlock · level recorded 2026-09-22 b4d2964f508734c3 · report
fixture: returned SoftComp · level recorded 2026-09-22 7bd366ee66ea9c84 · report
fixture: returned SoftSplit · level recorded 2026-09-22 f13da9bfe053deb9 · report
fixture: returned SparseWindowAttention · level recorded 2026-09-22 3d45197997a22569 · report
fixture: signature flow_fw:Tensor/4/float, flow_bw:Tensor/4/float, alpha1:float/0/float, alpha2:float/0/float · returned Tensor · level recorded 2026-09-22 6dba82ef5e1d54be · report
fixture: signature x:Tensor/4/float, flow:Tensor/4/float, interpolation:str/0/str, padding_mode:str/0/str, align_corners:bool/0/bool · returned Tensor · level recorded 2026-09-22 51f95e8de4711196 · report
fixture: signature negative_slope:float/0/float, inplace:bool/0/bool · returned function · level recorded 2026-09-22 f2b2ee9daa61fd8a · report
fixture: returned function · level recorded 2026-09-22 census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket pytorchcv/models/propainter.py ran; the contract check could not run (RAISES) MIT (permissive) ·
fe92e2df4330e9ac · report
fixture: signature x:Tensor/5/float, window_size:tuple/1/int, num_heads:int/0/int · returned Tensor · level recorded 2026-09-22 1949ce12af406baf · report
40b9d090f9b5f53d · report
5c1215e3af8ca9ac · report
f84760164d9dbd1e · report
0abf43743d60bcbc · report
3e21907371a62b09 · report
c9aea7b11102c2aa · report
Method as code Syntology
What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another.
2 assemblies on this paper: 2 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).
2 sibling assemblies for ProPainter (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.
harvested Assembly for ProPainter (graph method: ProPainter). 18 bricks; individually on a fixture: 12 ran (of which 7 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 4 where Syntology's instrument failed) · 6 unverified. The assembly has not been run end-to-end.
one of 2 siblings
- InpaintGenerator entry · level 0 · unverified ·
7b9c10bbcaf37395 - BaseNetwork helper · level 1 · ran ·
6a1c3869290d0d57 - BidirectionalPropagation model · level 0 · unverified ·
a7a87d9746b62c93 - Encoder model · level 1 · ran ·
53a5eb8ddb64103e - SoftComp model · level 0 · unverified ·
ae3ca4ef5eac8aec - SoftSplit model · level 1 · ran ·
9ff5a33dc359eb58 - TemporalSparseTransformerBlock helper · level 0 · unverified ·
fa4fe301319a48a2 - deconv helper · level 1 · ran ·
e73d218a85e7f736 - DeformableAlignment model · level 1 · ran ·
4cdb3a50bc7f846f - fbConsistencyCheck helper · level 1 · ran · our draft was wrong ·
e34f0d309218c4cb - flow_warp helper · level 1 · ran · our draft was wrong ·
0724aee9bf22b8ca - TemporalSparseTransformer model · level 0 · unverified ·
0355e712ec5750ab - ModulatedDeformConv2d helper · level 1 · ran ·
cbeb924f2247e025 - constant_init helper · level 0 · unverified ·
ae01e24194ffa4ed - length_sq helper · level 1 · ran · our draft was wrong ·
99919ce37333a615 - FusionFeedForward helper · level 1 · ran ·
83d5cf110edb5af1 - SparseWindowAttention helper · level 1 · ran ·
c4696b981f65032e - window_partition helper · level 1 · ran; the contract check could not run (RAISES) ·
c384ddd8232a9399
harvested Assembly for ProPainter (graph method: ProPainter). 30 bricks; individually on a fixture: 21 ran (of which 7 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 0 violated, 10 with no contract checked; 11 where Syntology's instrument failed) · 9 unverified. The assembly has not been run end-to-end.
one of 2 siblings
- ProPainter entry · level 0 · unverified ·
fc85da2c0e05c2ba - BidirectionalPropagation helper · level 0 · unverified ·
1949ce12af406baf - Decoder model · level 0 · unverified ·
40b9d090f9b5f53d - Encoder model · level 0 · unverified ·
5c1215e3af8ca9ac - SoftComp helper · level 1 · ran ·
b4d2964f508734c3 - SoftSplit helper · level 1 · ran ·
7bd366ee66ea9c84 - TemporalSparseTransformerBlock model · level 0 · unverified ·
c9aea7b11102c2aa - lambda_leakyrelu helper · level 1 · ran · our draft was wrong ·
51f95e8de4711196 - lambda_tanh helper · level 1 · ran ·
f2b2ee9daa61fd8a - ResBlock helper · level 0 · unverified ·
8f9a47ae470ac10f - SecondOrderDeformableAlignment helper · level 0 · unverified ·
0abf43743d60bcbc - fb_consistency_check helper · level 1 · ran · our draft was wrong ·
3d45197997a22569 - flow_warp helper · level 1 · ran · our draft was wrong ·
6dba82ef5e1d54be - PPDecoderUnit model · level 0 · unverified ·
f84760164d9dbd1e - conv3x3_block helper · level 1 · ran · our draft was wrong ·
9dc286722d0492ff - conv3x3 helper · level 1 · ran · our draft was wrong ·
7f388bc44b6fceae - TemporalSparseTransformer model · level 0 · unverified ·
3e21907371a62b09 - lambda_batchnorm2d helper · level 1 · ran · our draft was wrong ·
0aa50f8a83f710b9 - lambda_relu helper · level 1 · ran · our draft was wrong ·
71d18c7938254ec3 - length_sq helper · level 1 · ran · our draft was wrong ·
99919ce37333a615 - InterpolationBlock helper · level 1 · ran ·
2fd5dd72e6bde9a9 - ConvBlock helper · level 1 · ran ·
c69c35f159717fe0 - FusionFeedForward model · level 1 · ran ·
83365c19252ece6f - SparseWindowAttention model · level 1 · ran ·
f13da9bfe053deb9 - create_activation_layer helper · level 1 · ran · our draft was wrong ·
6ab307b869a70f42 - create_normalization_layer helper · level 1 · ran · our draft was wrong ·
caa1617c550d8b47 - window_partition helper · level 1 · ran; the contract check could not run (RAISES) ·
fe92e2df4330e9ac - HSigmoid helper · level 1 · ran ·
fc85bca2cba2a282 - HSwish helper · level 1 · ran ·
861080d22f42f5ab - Swish helper · level 1 · ran ·
6152b289f1a1e079
Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2309.03897") returns these assemblies with each brick's role and wiring (how to connect).
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Video Inpainting | HQVI (240p) | ProPainter | LPIPS | 0.0388 | #2 of 7 | Archive leaderboard | report |
| Video Inpainting | HQVI (240p) | ProPainter | PSNR | 30.62 | #2 of 7 | Archive leaderboard | report |
| Video Inpainting | HQVI (240p) | ProPainter | SSIM | 0.9413 | #2 of 7 | Archive leaderboard | report |
| Video Inpainting | HQVI (240p) | ProPainter | VFID | 0.2128 | #2 of 7 | Archive leaderboard | report |
| Video Inpainting | HQVI (480p) | ProPainter | LPIPS | 0.0457 | #4 of 4 | Archive leaderboard | report |
| Video Inpainting | HQVI (480p) | ProPainter | PSNR | 30.69 | #4 of 4 | Archive leaderboard | report |
| Video Inpainting | HQVI (480p) | ProPainter | SSIM | 0.9414 | #4 of 4 | Archive leaderboard | report |
| Video Inpainting | HQVI (480p) | ProPainter | VFID | 0.0478 | #4 of 4 | Archive leaderboard | report |
| Video Inpainting | YouTube-VOS 2018 | ProPainter | Ewarp | - | #1 of 10 | Archive leaderboard | report |
| Video Inpainting | YouTube-VOS 2018 | ProPainter | PSNR | 34.43 | #1 of 10 | Archive leaderboard | report |
| Video Inpainting | YouTube-VOS 2018 | ProPainter | SSIM | 0.9735 | #1 of 10 | Archive leaderboard | report |
| Video Inpainting | YouTube-VOS 2018 | ProPainter | VFID | 0.042 | #1 of 10 | 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
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