Papers › Swin Transformer V2: Scaling Up Capacity and Resolution
Swin Transformer V2: Scaling Up Capacity and Resolution
Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, Baining Guo
Large-scale NLP models have been shown to significantly improve the performance on language tasks with no signs of saturation. They also demonstrate amazing few-shot capabilities like that of human beings. This paper aims to explore large-scale models in computer vision. We tackle three major issues in training and application of large vision models, including training instability, resolution gaps between pre-training and fine-tuning, and hunger on labelled data. Three main techniques are proposed: 1) a residual-post-norm method combined with cosine attention to improve training stability; 2) A log-spaced continuous position bias method to effectively transfer models pre-trained using low-resolution images to downstream tasks with high-resolution inputs; 3) A self-supervised pre-training method, SimMIM, to reduce the needs of vast labeled images. Through these techniques, this paper successfully trained a 3 billion-parameter Swin Transformer V2 model, which is the largest dense vision model to date, and makes it capable of training with images of up to 1,536×1,536 resolution. It set new performance records on 4 representative vision tasks, including ImageNet-V2 image classification, COCO object detection, ADE20K semantic segmentation, and Kinetics-400 video action classification. Also note our training is much more efficient than that in Google's billion-level visual models, which consumes 40 times less labelled data and 40 times less training time. Code is available at https://github.com/microsoft/Swin-Transformer.
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="2111.09883")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2111.09883", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2111.09883")
Code
No sample from the archive's official repository (microsoft/Swin-Transformer) is linked to this paper; the 30 samples below come from 4 community repositories. Syntology harvested code from microsoft/Swin-Transformer for another paper.
microsoft/Swin-Transformer: Syntology's graph links these samples to Swin Transformer: Hierarchical Vision Transformer using Shifted Windows (arXiv:2103.14030; that paper's own run record: 123 ran (of which 45 constructed an object rather than computing a result; 79 with no instrument failure: 6 honoured, 3 violated, 70 with no contract checked; 44 where Syntology's instrument failed) · 84 unverified); nothing checks that they implement this paper's method.
Syntology 30 code samples harvested from 4 repositories linked to this paper: 17 ran (of which 0 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 3 violated, 11 with no contract checked; 3 where Syntology's instrument failed) · 13 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; failures of Syntology's instrument, not of the code.
By repository: community (archive-listed): 30 samples from 4 repositories, 17 ran (14 with no instrument failure; 3 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.
No sample is linked to this paper from 15 of its repositories that Syntology harvested for other papers:
- microsoft/Swin-Transformer, official repository: harvested for another paper, listed under the repository
- Burf/SwinTransformer-Tensorflow2, community (archive-listed): harvested for another paper, listed under the repository
- ChristophReich1996/Swin-Transformer-V2, community (archive-listed): harvested for another paper, listed under the repository
- impiga/plain-detr, community (archive-listed): harvested for another paper, listed under the repository
- isaaccorley/hydro-foundation-model, community (archive-listed): harvested for another paper, listed under the repository
- lucidrains/flash-cosine-sim-attention, community (archive-listed): harvested for 2 other papers, listed under the repository
- rami0205/ngramswin, community (archive-listed): harvested for 2 other papers, listed under the repository
- rwightman/pytorch-image-models, community (archive-listed): harvested for another paper, listed under the repository
- Burf/tfdetection, community (archive-listed): harvested for 4 other papers, listed under the repository
- MindCode-4/code-5, community (archive-listed): harvested for another paper, listed under the repository
- PaddlePaddle/PaddleDetection, community (archive-listed): harvested for another paper, listed under the repository
- leondgarse/keras_cv_attention_models, community (archive-listed): harvested for another paper, listed under the repository
- pwc-1/Paper-10, community (archive-listed): harvested for another paper, listed under the repository
- shkarupa-alex/tfswin, community (archive-listed): harvested for another paper, listed under the repository
- towhee-io/towhee, community (archive-listed): harvested for 12 other papers, listed under the repository
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. 1 of the 17 samples that ran has no recorded date.
23 repositories listed; official and paper-mentioned ones first.
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.
A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.
Code Syntology ran Syntology
30 samples harvested: 17 ran (of which 0 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 3 violated, 11 with no contract checked; 3 where Syntology's instrument failed) · 13 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.
Licence: 4 of the 30 samples are pointer only, meaning Syntology does not serve that copy's text. For 4 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.
Harvested from 4 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”. “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.
1f7c550e65142ef8 · report
fixture: signature val:int/0/int, depth:int/0/int · returned tuple · level recorded 2026-09-22 424012cb37b31172 · report
fixture: signature val:NoneType/0/?, d:callable/0/? · returned int · level recorded 2026-09-22 census: 5 implementations from 103 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 1, kind int, dtype int64 pd_x_transformers/autoregressive_wrapper.py ran · violated contract Apache-2.0 (permissive) here; the stored copy is pointer only ·
aa5486a3650902d8 · report
fixture: signature val:Tensor/1/int · returned bool · level recorded, date not recorded eaade2de97f347bc · report
fixture: signature group:list/1/int · returned bool · level recorded 2026-09-24 29b5922a6530d4e2 · report
942acecd8aa049a5 · report
80c04c4768d37ac0 · report
fixture: signature image_shape:tuple/1/int, lam:float/0/float, minmax:NoneType/0/?, correct_lam:bool/0/bool, count:NoneType/0/? · returned tuple · level recorded 2026-09-24 043c2f16bd404117 · report
fixture: signature filename:str/0/str, logger_name:str/0/str · returned Logger · level recorded 2026-09-24 484edb411de743ae · report
fixture: signature image_shape:tuple/1/int, lam:float/0/float, count:int/0/int · returned tuple · level recorded 2026-09-24 b27df77c248443b3 · report
fixture: signature image_shape:tuple/1/int, minmax:tuple/1/float, count:int/0/int · returned tuple · level recorded 2026-09-24 e0254103b7a7465f · report
fixture: signature image:Image/0/?, magnitude:float/0/float, fillcolor:tuple/1/int · returned Image · level recorded 2026-09-24 00da552067622c6f · report
fixture: signature image:Image/0/?, magnitude:float/0/float, fillcolor:tuple/1/int · returned Image · level recorded 2026-09-24 5931b70ac74f0d83 · report
fixture: signature image:Image/0/?, magnitude:float/0/float, fillcolor:tuple/1/int · returned Image · level recorded 2026-09-24 1e4ddb470d424318 · report
fixture: signature config:MockConfig/0/?, args:MockArgs/0/? · returned MockConfig · level recorded 2026-09-24 e14035f3e23f1d93 · report
21421188135068e9 · report
fixture: signature batch_size:int/0/int, samples_each_class:int/0/int · returned ndarray · level recorded 2026-09-24 f262aa35d9f2b50f · report
fixture: signature name:str/0/str, names:list/1/str, thresh:float/0/float, topk:int/0/int · returned list · level recorded 2026-09-24 census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket ppcls/arch/backbone/model_zoo/swin_transformer.py unverified Apache-2.0 (permissive) ·
0fd983e99eaff69e · report
1a004818e36b501f · report
177a4abf40724f6e · report
fixture: signature st:str/0/str · returned str · level recorded 2026-09-25 ffc9d7cced6f71fb · report
fixture: signature argv:list/1/str · returned list · level recorded 2026-09-25 a99f170ebd682af5 · report
fixture: signature st:str/0/str · returned str · level recorded 2026-09-25 census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket src/models/swin_transformer.py unverified Apache-2.0 (permissive) ·
73879546dbb5b1b1 · report
1129d77773d2fc84 · report
db165c23fedab0fa · report
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Action Classification | Kinetics-400 | Video-SwinV2-G (ImageNet-22k and external 70M pretrain) | Acc@1 | 86.8 | #44 of 207 | Archive leaderboard | report |
| Image Classification | ImageNet | SwinV2-G | Number of params | 3000M | #7 of 1060 | Archive leaderboard | report |
| Image Classification | ImageNet | SwinV2-G | Top 1 Accuracy | 90.17% | #7 of 1060 | Archive leaderboard | report |
| Image Classification | ImageNet | SwinV2-B | Number of params | 88M | #106 of 1060 | Archive leaderboard | report |
| Image Classification | ImageNet | SwinV2-B | Top 1 Accuracy | 87.1% | #106 of 1060 | Archive leaderboard | report |
| Image Classification | ImageNet V2 | SwinV2-G | Top 1 Accuracy | 84.00% | #4 of 33 | Archive leaderboard | report |
| Image Classification | ImageNet V2 | SwinV2-B | Top 1 Accuracy | 78.08 | #13 of 33 | Archive leaderboard | report |
| Instance Segmentation | COCO minival | SwinV2-G (HTC++) | mask AP | 53.7 | #9 of 93 | Archive leaderboard | report |
| Instance Segmentation | COCO test-dev | SwinV2-G (HTC++) | mask AP | 54.4 | #10 of 112 | Archive leaderboard | report |
| Object Detection | COCO minival | SwinV2-G (HTC++) | box AP | 62.5 | #15 of 220 | Archive leaderboard | report |
| Object Detection | COCO test-dev | SwinV2-G (HTC++) | Params (M) | 3000 | #18 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | SwinV2-G (HTC++) | box mAP | 63.1 | #18 of 225 | Archive leaderboard | report |
| Semantic Segmentation | ADE20K | SwinV2-G(UperNet) | Validation mIoU | 59.9 | #15 of 235 | Archive leaderboard | report |
| Semantic Segmentation | ADE20K | SwinV2-G-HTC++ Liu et al. ([2021a]) | Validation mIoU | 53.7 | #75 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
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