Papers › FuseFormer: Fusing Fine-Grained Information in Transformers for Video Inpainting

FuseFormer: Fusing Fine-Grained Information in Transformers for Video Inpainting

7 Sep 2021ICCV 2021 10arXiv:2109.02974archive 2025-07-28

Rui Liu, Hanming Deng, Yangyi Huang, Xiaoyu Shi, Lewei Lu, Wenxiu Sun, Xiaogang Wang, Jifeng Dai, Hongsheng Li

Transformer, as a strong and flexible architecture for modelling long-range relations, has been widely explored in vision tasks. However, when used in video inpainting that requires fine-grained representation, existed method still suffers from yielding blurry edges in detail due to the hard patch splitting. Here we aim to tackle this problem by proposing FuseFormer, a Transformer model designed for video inpainting via fine-grained feature fusion based on novel Soft Split and Soft Composition operations. The soft split divides feature map into many patches with given overlapping interval. On the contrary, the soft composition operates by stitching different patches into a whole feature map where pixels in overlapping regions are summed up. These two modules are first used in tokenization before Transformer layers and de-tokenization after Transformer layers, for effective mapping between tokens and features. Therefore, sub-patch level information interaction is enabled for more effective feature propagation between neighboring patches, resulting in synthesizing vivid content for hole regions in videos. Moreover, in FuseFormer, we elaborately insert the soft composition and soft split into the feed-forward network, enabling the 1D linear layers to have the capability of modelling 2D structure. And, the sub-patch level feature fusion ability is further enhanced. In both quantitative and qualitative evaluations, our proposed FuseFormer surpasses state-of-the-art methods. We also conduct detailed analysis to examine its superiority.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2109.02974")

Code

Syntology Ran 10 of 11 code samples harvested from 1 repository linked to this paper; 1 has no recorded run. Of those that ran: 10 ran with no contract checked.

By repository: official repository: 11 samples from 1 repository, 10 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

ruiliu-ai/fuseformer officialmentioned in papermentioned on GitHubpytorch 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

11 samples harvested; 10 ran; 0 honoured the contract we drafted; 1 has no recorded run. Read from Syntology's graph 2026-09-25; that is when this build read the record, not when the samples ran.

10ran
1unverified

Licence: 11 of the 11 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 ruiliu-ai/FuseFormer. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

AddPosEmb ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · 974fc45de281481c · report
Attention ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic fingerprinted no licence file found · pointer only · a702c7ce4cea9802 · report
BaseNetwork ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · 6a1c3869290d0d57 · report
Encoder ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran fingerprinted no licence file found · pointer only · 5231e4794aac485a · report
FusionFeedForward ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · 3d843c10ed59c635 · report
MultiHeadedAttention ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · 634deecea407ff83 · report
SoftComp ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · b99e0f5e4538ded3 · report
SoftSplit ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · 2e02b61d792f9198 · report
TransformerBlock ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic no licence file found · pointer only · afa615aa2f60ae16 · report
deconv ruiliu-ai/FuseFormer/model/fuseformer.py official repository ran · metamorphic tier: deterministic fingerprinted no licence file found · pointer only · e73d218a85e7f736 · report
InpaintGenerator ruiliu-ai/FuseFormer/model/fuseformer.py official repository unverified no licence file found · pointer only · dbafafc1baa0fc1c · report

Tasks

Seeing Beyond the VisibleVideo Inpainting

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Seeing Beyond the Visible KITTI360-EX FuseFormer Average PSNR 18.91 #4 of 7 Archive leaderboard report
Video Inpainting DAVIS FuseFormer Ewarp 0.1362 #3 of 11 Archive leaderboard report
Video Inpainting DAVIS FuseFormer PSNR 32.54 #3 of 11 Archive leaderboard report
Video Inpainting DAVIS FuseFormer SSIM 0.9700 #3 of 11 Archive leaderboard report
Video Inpainting DAVIS FuseFormer VFID 0.138 #3 of 11 Archive leaderboard report
Video Inpainting HQVI (240p) FuseFormer LPIPS 0.0498 #6 of 7 Archive leaderboard report
Video Inpainting HQVI (240p) FuseFormer PSNR 29.92 #6 of 7 Archive leaderboard report
Video Inpainting HQVI (240p) FuseFormer SSIM 0.9365 #6 of 7 Archive leaderboard report
Video Inpainting HQVI (240p) FuseFormer VFID 0.2727 #6 of 7 Archive leaderboard report
Video Inpainting YouTube-VOS 2018 FuseFormer Ewarp 0.0900 #4 of 10 Archive leaderboard report
Video Inpainting YouTube-VOS 2018 FuseFormer PSNR 33.29 #4 of 10 Archive leaderboard report
Video Inpainting YouTube-VOS 2018 FuseFormer SSIM 0.9681 #4 of 10 Archive leaderboard report
Video Inpainting YouTube-VOS 2018 FuseFormer VFID 0.053 #4 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

Introduced by this paper: FuseFormer Block, Soft Split and Soft Composition

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutFuseFormerFuseFormer BlockInpaintingLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoft Split and Soft CompositionSoftmaxTransformer

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