Papers › DeVIS: Making Deformable Transformers Work for Video Instance Segmentation

DeVIS: Making Deformable Transformers Work for Video Instance Segmentation

22 Jul 2022arXiv:2207.11103archive 2025-07-28

Adrià Caelles, Tim Meinhardt, Guillem Brasó, Laura Leal-Taixé

Video Instance Segmentation (VIS) jointly tackles multi-object detection, tracking, and segmentation in video sequences. In the past, VIS methods mirrored the fragmentation of these subtasks in their architectural design, hence missing out on a joint solution. Transformers recently allowed to cast the entire VIS task as a single set-prediction problem. Nevertheless, the quadratic complexity of existing Transformer-based methods requires long training times, high memory requirements, and processing of low-single-scale feature maps. Deformable attention provides a more efficient alternative but its application to the temporal domain or the segmentation task have not yet been explored. In this work, we present Deformable VIS (DeVIS), a VIS method which capitalizes on the efficiency and performance of deformable Transformers. To reason about all VIS subtasks jointly over multiple frames, we present temporal multi-scale deformable attention with instance-aware object queries. We further introduce a new image and video instance mask head with multi-scale features, and perform near-online video processing with multi-cue clip tracking. DeVIS reduces memory as well as training time requirements, and achieves state-of-the-art results on the YouTube-VIS 2021, as well as the challenging OVIS dataset. Code is available at https://github.com/acaelles97/DeVIS.

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 11 code samples harvested from 1 repository linked to this paper: 9 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 1 honoured, 0 violated, 7 with no contract checked; 1 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 11 samples from 1 repository, 9 ran (8 with no instrument failure; 1 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 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

acaelles97/devis officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 11 samples, 9 ran (1 instrument-failed) · 2 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

11 samples harvested: 9 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 1 honoured, 0 violated, 7 with no contract checked; 1 where Syntology's instrument failed) · 2 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.

1ran · honoured contract
1ran · our draft was wrong
7ran
2unverified

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 acaelles97/devis. “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.

acaelles97/devis official repository 11 samples, 9 ran (1 instrument-failed) · 2 unverified
build_3D_conv_head src/models/devis_segmentation.py ran Apache-2.0 recorded; this copy not marked cleared · pointer only · 894e7d5a12a809b7 · report fixture: signature mask_head_hidden_size:int/0/int · returned Sequential · level recorded 2026-09-25
convert_to_xywh in 26 papers src/datasets/coco_eval.py ran · honoured contract fingerprinted Apache-2.0 recorded; this copy not marked cleared · pointer only · f31a58bf6457ced5 · report fixture: signature boxes:Tensor/2/float · returned Tensor · level recorded 2026-09-22
crop in 99 papers
census: 5 implementations from 5 papers, 4 distinct outputs across 4 buckets, one shared input per bucket
src/datasets/coco_transforms.py ran Apache-2.0 recorded; this copy not marked cleared · pointer only · 8ea3bd3cae92680d · report fixture: signature image:Tensor/3/float, target:dict/0/?, region:tuple/1/int · returned tuple · level recorded 2026-09-25
evaluate in 514 papers src/datasets/coco_eval.py ran · our draft was wrong Apache-2.0 recorded; this copy not marked cleared · pointer only · fe0ddcc2d420c9a0 · report fixture: signature self:MockCOCOEval/0/? · returned tuple · level recorded 2026-09-22
get_size_with_aspect_ratio src/datasets/coco_transforms.py ran Apache-2.0 recorded; this copy not marked cleared · pointer only · 2354640608431f61 · report fixture: signature image_size:tuple/1/int, size:int/0/int, max_size:int/0/int · returned tuple · level recorded 2026-09-25
hflip in 29 papers src/datasets/coco_transforms.py ran Apache-2.0 recorded; this copy not marked cleared · pointer only · 344591b983c871de · report fixture: signature image:Tensor/3/int, target:dict/0/? · returned tuple · level recorded 2026-09-25
merge in 85 papers src/datasets/coco_eval.py ran Apache-2.0 recorded; this copy not marked cleared · pointer only · eb4edcefd813ed1f · report fixture: returned tuple · level recorded 2026-09-25
process_centroid visualize_dataset.py ran Apache-2.0 recorded; this copy not marked cleared · pointer only · c1af59f873dcaf84 · report fixture: signature centroids:Tensor/2/float, tgt_size:tuple/1/int · returned Tensor · level recorded 2026-09-25
un_normalize visualize_dataset.py ran fingerprinted Apache-2.0 recorded; this copy not marked cleared · pointer only · 96420c6020086b5c · report fixture: signature tensor:Tensor/3/float · returned Tensor · level recorded 2026-09-25
make_coco_train_transforms src/datasets/coco.py unverified Apache-2.0 recorded; this copy not marked cleared · pointer only · 9fc36974a286ba3c · report
make_coco_val_transforms src/datasets/coco.py unverified Apache-2.0 recorded; this copy not marked cleared · pointer only · 08f78c9f2a7c5d05 · report

Tasks

Instance SegmentationObject DetectionSegmentationSemantic SegmentationVideo Instance Segmentationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Video Instance Segmentation OVIS validation DeVIS (Swin-L) AP50 59.3 #25 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (Swin-L) AP75 38.3 #25 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (Swin-L) AR1 16.6 #25 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (Swin-L) AR10 39.8 #25 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (Swin-L) mask AP 35.5 #25 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (ResNet-50) AP50 47.6 #33 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (ResNet-50) AP75 20.8 #33 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (ResNet-50) AR1 12.0 #33 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (ResNet-50) AR10 28.9 #33 of 44 Archive leaderboard report
Video Instance Segmentation OVIS validation DeVIS (ResNet-50) mask AP 23.7 #33 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (Swin-L) AP50 77.7 #17 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (Swin-L) AP75 59.8 #17 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (Swin-L) AR1 43.8 #17 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (Swin-L) AR10 57.8 #17 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (Swin-L) mask AP 54.4 #17 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (ResNet-50) AP50 66.8 #24 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (ResNet-50) AP75 46.6 #24 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (ResNet-50) AR1 38.0 #24 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (ResNet-50) AR10 50.1 #24 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS 2021 DeVIS (ResNet-50) mask AP 43.1 #24 of 26 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (Swin-L) AP50 80.8 #10 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (Swin-L) AP75 66.3 #10 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (Swin-L) AR1 50.8 #10 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (Swin-L) AR10 61.0 #10 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (Swin-L) mask AP 57.1 #10 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (ResNet-50) AP50 66.7 #23 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (ResNet-50) AP75 48.6 #23 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (ResNet-50) AR1 42.4 #23 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (ResNet-50) AR10 51.6 #23 of 44 Archive leaderboard report
Video Instance Segmentation YouTube-VIS validation DeVIS (ResNet-50) mask AP 44.4 #23 of 44 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

CLIP

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