Papers › Vision Transformer with Deformable Attention

Vision Transformer with Deformable Attention

3 Jan 2022CVPR 2022 1arXiv:2201.00520archive 2025-07-28

Zhuofan Xia, Xuran Pan, Shiji Song, Li Erran Li, Gao Huang

Transformers have recently shown superior performances on various vision tasks. The large, sometimes even global, receptive field endows Transformer models with higher representation power over their CNN counterparts. Nevertheless, simply enlarging receptive field also gives rise to several concerns. On the one hand, using dense attention e.g., in ViT, leads to excessive memory and computational cost, and features can be influenced by irrelevant parts which are beyond the region of interests. On the other hand, the sparse attention adopted in PVT or Swin Transformer is data agnostic and may limit the ability to model long range relations. To mitigate these issues, we propose a novel deformable self-attention module, where the positions of key and value pairs in self-attention are selected in a data-dependent way. This flexible scheme enables the self-attention module to focus on relevant regions and capture more informative features. On this basis, we present Deformable Attention Transformer, a general backbone model with deformable attention for both image classification and dense prediction tasks. Extensive experiments show that our models achieve consistently improved results on comprehensive benchmarks. Code is available at https://github.com/LeapLabTHU/DAT.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

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 10 code samples harvested from 2 repositories linked to this paper: 8 ran (of which 4 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 4 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 2 samples from 1 repository, 2 ran (2 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 8 samples from 1 repository, 6 ran (2 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 on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

leaplabthu/dat officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 2 samples, 2 ran (0 instrument-failed) · 0 unverified report
ChristophReich1996/Swin-Transformer-V2 mentioned on GitHubpytorchSyntology: 8 samples, 6 ran (4 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

10 samples harvested: 8 ran (of which 4 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 4 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.

3ran · our draft was wrong
1ran; the contract check could not run
4ran
2unverified

Licence: 0 of the 10 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 3 samples 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.

leaplabthu/dat official repository 2 samples, 2 ran (0 instrument-failed) · 0 unverified
DAttentionBaseline models/dat_blocks.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · 7cba8d74ab462cab · report fixture: returned DAttentionBaseline · level recorded 2026-09-22
LayerNormProxy models/dat_blocks.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 5ebf85fee0f0738f · report fixture: returned LayerNormProxy · level recorded 2026-09-22
ChristophReich1996/Swin-Transformer-V2 community (archive-listed) 8 samples, 6 ran (4 instrument-failed) · 2 unverified
FeedForward in 198 papers swin_transformer_v2/model_parts.py ran MIT (permissive) · 2aa71e543f03992a · report fixture: returned FeedForward · level recorded 2026-09-22
WindowMultiHeadAttention swin_transformer_v2/model_parts.py ran MIT (permissive) · 5716ef7256430436 · report fixture: returned WindowMultiHeadAttention · level recorded 2026-09-22
bchw_to_bhwc swin_transformer_v2/model_parts.py ran · our draft was wrong fingerprinted MIT (permissive) · f76dfdc07001b5e3 · report fixture: signature input:Tensor/4/float · returned Tensor · level recorded 2026-09-22
bhwc_to_bchw swin_transformer_v2/model_parts.py ran · our draft was wrong fingerprinted MIT (permissive) · 8c3d7c2fece98766 · report fixture: signature input:Tensor/4/float · returned Tensor · level recorded 2026-09-22
fold in 10 papers swin_transformer_v2/model_parts.py ran · our draft was wrong MIT (permissive) · ae92018684b30199 · report fixture: signature input:Tensor/4/float, window_size:int/0/int, height:int/0/int, width:int/0/int · returned Tensor · level recorded 2026-09-22
unfold in 7 papers swin_transformer_v2/model_parts.py ran; the contract check could not run (RAISES) fingerprinted MIT (permissive) · 31576ed450a2bd34 · report fixture: signature input:Tensor/4/float, window_size:int/0/int · returned Tensor · level recorded 2026-09-22
DeformableSwinTransformerBlock swin_transformer_v2/model_parts.py unverified MIT (permissive) · 6fba747e5af83c5f · report
SwinTransformerBlock in 35 papers swin_transformer_v2/model_parts.py unverified MIT (permissive) · 3b8f635fca41be47 · 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 Deformable Attention Transformer (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for Deformable Attention Transformer (graph method: Deformable Attention Transformer). 2 bricks; individually on a fixture: 2 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; every one of the 2 samples that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 2 siblings

  1. DAttentionBaseline entry · level 1 · ran · 7cba8d74ab462cab
  2. LayerNormProxy helper · level 1 · ran · 5ebf85fee0f0738f

harvested Assembly for Deformable Attention Transformer (graph method: Deformable Attention Transformer). 8 bricks; individually on a fixture: 6 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 4 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. DeformableSwinTransformerBlock entry · level 0 · unverified · 6fba747e5af83c5f
  2. SwinTransformerBlock helper · level 0 · unverified · 3b8f635fca41be47
  3. FeedForward helper · level 1 · ran · 2aa71e543f03992a
  4. WindowMultiHeadAttention helper · level 1 · ran · 5716ef7256430436
  5. bchw_to_bhwc helper · level 1 · ran · our draft was wrong · f76dfdc07001b5e3
  6. bhwc_to_bchw helper · level 1 · ran · our draft was wrong · 8c3d7c2fece98766
  7. fold helper · level 1 · ran · our draft was wrong · ae92018684b30199
  8. unfold helper · level 1 · ran; the contract check could not run (RAISES) · 31576ed450a2bd34

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="2201.00520") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Image ClassificationObject DetectionSemantic Segmentationimage-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ImageNet DAT-B (384 res, IN-1K only) GFLOPs 49.8 #289 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-B (384 res, IN-1K only) Number of params 88M #289 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-B (384 res, IN-1K only) Top 1 Accuracy 84.8% #289 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-S GFLOPs 9.0 #396 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-S Number of params 50M #396 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-S Top 1 Accuracy 83.7% #396 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-T GFLOPs 4.6 #588 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-T Number of params 29M #588 of 1060 Archive leaderboard report
Image Classification ImageNet DAT-T Top 1 Accuracy 82.0% #588 of 1060 Archive leaderboard report
Object Detection COCO test-dev DAT-S (RetinaNet) AP50 69.6 #112 of 225 Archive leaderboard report
Object Detection COCO test-dev DAT-S (RetinaNet) AP75 51.2 #112 of 225 Archive leaderboard report
Object Detection COCO test-dev DAT-S (RetinaNet) APL 63.4 #112 of 225 Archive leaderboard report
Object Detection COCO test-dev DAT-S (RetinaNet) APM 51.8 #112 of 225 Archive leaderboard report
Object Detection COCO test-dev DAT-S (RetinaNet) APS 32.3 #112 of 225 Archive leaderboard report
Object Detection COCO test-dev DAT-S (RetinaNet) box mAP 47.9 #112 of 225 Archive leaderboard report
Semantic Segmentation ADE20K DAT-B (UperNet) Params (M) 121 #133 of 235 Archive leaderboard report
Semantic Segmentation ADE20K DAT-B (UperNet) Validation mIoU 49.38 #133 of 235 Archive leaderboard report
Semantic Segmentation ADE20K DAT-S (UperNet) Params (M) 81 #148 of 235 Archive leaderboard report
Semantic Segmentation ADE20K DAT-S (UperNet) Validation mIoU 48.31 #148 of 235 Archive leaderboard report
Semantic Segmentation ADE20K DAT-T (UperNet) Params (M) 60 #190 of 235 Archive leaderboard report
Semantic Segmentation ADE20K DAT-T (UperNet) Validation mIoU 45.54 #190 of 235 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 AttentionPVTPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxSpatial-Reduction AttentionStochastic DepthSwin TransformerTransformer

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