Papers › HRFormer: High-Resolution Transformer for Dense Prediction

HRFormer: High-Resolution Transformer for Dense Prediction

18 Oct 2021arXiv:2110.09408archive 2025-07-28

Yuhui Yuan, Rao Fu, Lang Huang, WeiHong Lin, Chao Zhang, Xilin Chen, Jingdong Wang

We present a High-Resolution Transformer (HRFormer) that learns high-resolution representations for dense prediction tasks, in contrast to the original Vision Transformer that produces low-resolution representations and has high memory and computational cost. We take advantage of the multi-resolution parallel design introduced in high-resolution convolutional networks (HRNet), along with local-window self-attention that performs self-attention over small non-overlapping image windows, for improving the memory and computation efficiency. In addition, we introduce a convolution into the FFN to exchange information across the disconnected image windows. We demonstrate the effectiveness of the High-Resolution Transformer on both human pose estimation and semantic segmentation tasks, e.g., HRFormer outperforms Swin transformer by 1.3 AP on COCO pose estimation with 50% fewer parameters and 30% fewer FLOPs. Code is available at: https://github.com/HRNet/HRFormer.

PaperPDFCodeCode 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 11 code samples harvested from 1 repository linked to this paper: 7 ran (of which 7 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified; every one of the 7 samples that ran constructed an object rather than computing a result. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code; here there are none.

By repository: official repository: 11 samples from 1 repository, 7 ran (7 with no instrument failure; 0 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.

HRNet/HRFormer officialmentioned in papermentioned on GitHubpytorchSyntology: 11 samples, 7 ran (0 instrument-failed) · 4 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: 7 ran (of which 7 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified; every one of the 7 samples that ran constructed an object rather than computing a result. “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.

7ran
4unverified

Licence: 0 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 HRNet/HRFormer. “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.

HRNet/HRFormer official repository 11 samples, 7 ran (0 instrument-failed) · 4 unverified
Bottleneck in 98 papers cls/models/hrt.py ran · metamorphic tier: deterministic MIT (permissive) · 70a3b60f25883442 · report fixture: returned Bottleneck · level recorded 2026-09-22
BottleneckDWP cls/models/hrt.py ran · metamorphic tier: deterministic fingerprinted MIT (permissive) · 25d275a698577c77 · report fixture: returned BottleneckDWP · level recorded 2026-09-22
HighResolutionTransformerModule cls/models/hrt.py ran · metamorphic tier: deterministic MIT (permissive) · dc476e7863dfa60a · report fixture: returned HighResolutionTransformerModule · level recorded 2026-09-22
LocalPermuteModule cls/models/hrt.py ran MIT (permissive) · f3ae82b7f0fa5762 · report fixture: returned LocalPermuteModule · level recorded 2026-09-22
MlpDWBN cls/models/hrt.py ran · metamorphic tier: deterministic MIT (permissive) · 7511c8343dacb63e · report fixture: returned MlpDWBN · level recorded 2026-09-22
MultiheadAttention in 44 papers cls/models/hrt.py ran · metamorphic tier: deterministic MIT (permissive) · 7f956321be67f123 · report fixture: returned MultiheadAttention · level recorded 2026-09-22
PadBlock cls/models/hrt.py ran MIT (permissive) · f1a2da4545324be4 · report fixture: returned PadBlock · level recorded 2026-09-22
GeneralTransformerBlock cls/models/hrt.py unverified MIT (permissive) · 174ded5efb8d90bb · report
HighResolutionTransformer cls/models/hrt.py unverified MIT (permissive) · a219e0495156d384 · report
MultiheadAttentionRPE cls/models/hrt.py unverified MIT (permissive) · 77bb6724a9d5ee58 · report
MultiheadISAAttention cls/models/hrt.py unverified MIT (permissive) · dd8ca007aa495f2f · 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.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for HRFormer (graph method: HRFormer). 11 bricks; individually on a fixture: 7 ran (of which 7 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified; every one of the 7 samples that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

  1. HighResolutionTransformer entry · level 0 · unverified · a219e0495156d384
  2. BottleneckDWP helper · level 1 · ran · 25d275a698577c77
  3. GeneralTransformerBlock model · level 0 · unverified · 174ded5efb8d90bb
  4. HighResolutionTransformerModule model · level 1 · ran · dc476e7863dfa60a
  5. MlpDWBN helper · level 1 · ran · 7511c8343dacb63e
  6. MultiheadISAAttention model · level 0 · unverified · dd8ca007aa495f2f
  7. LocalPermuteModule helper · level 1 · ran · f3ae82b7f0fa5762
  8. MultiheadAttentionRPE helper · level 0 · unverified · 77bb6724a9d5ee58
  9. PadBlock helper · level 1 · ran · f1a2da4545324be4
  10. MultiheadAttention model · level 1 · ran · 7f956321be67f123
  11. Bottleneck helper · level 1 · ran · 70a3b60f25883442

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="2110.09408") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Image ClassificationMulti-Person Pose EstimationPose EstimationPredictionSemantic SegmentationVocal Bursts Intensity Prediction

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ImageNet HRFormer-B GFLOPs 13.7 #503 of 1060 Archive leaderboard report
Image Classification ImageNet HRFormer-B Number of params 50.3M #503 of 1060 Archive leaderboard report
Image Classification ImageNet HRFormer-B Top 1 Accuracy 82.8% #503 of 1060 Archive leaderboard report
Image Classification ImageNet HRFormer-T GFLOPs 1.8 #827 of 1060 Archive leaderboard report
Image Classification ImageNet HRFormer-T Number of params 8.0M #827 of 1060 Archive leaderboard report
Image Classification ImageNet HRFormer-T Top 1 Accuracy 78.5% #827 of 1060 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose HRFormer-B AP Easy 80.0 #7 of 28 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose HRFormer-B AP Hard 62.4 #7 of 28 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose HRFormer-B AP Medium 73.5 #7 of 28 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose HRFormer-B mAP @0.5:0.95 72.4 #7 of 28 Archive leaderboard report
Multi-Person Pose Estimation OCHuman HRFormer-B AP50 81.4 #4 of 8 Archive leaderboard report
Multi-Person Pose Estimation OCHuman HRFormer-B AP75 67.1 #4 of 8 Archive leaderboard report
Multi-Person Pose Estimation OCHuman HRFormer-B Validation AP 62.1 #4 of 8 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-B) AP 34.4 #3 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-B) AP50 78.3 #3 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-B) AP75 24.8 #3 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-B) AR 38.7 #3 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-B) AR50 80.9 #3 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-S) AP 31.6 #7 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-S) AP75 20.9 #7 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-S) AR 35.8 #7 of 10 Archive leaderboard report
Pose Estimation AIC HRFormer (HRFomer-S) AR50 78.0 #7 of 10 Archive leaderboard report
Pose Estimation COCO test-dev HRFormer-B AP 76.2 #17 of 47 Archive leaderboard report
Pose Estimation COCO test-dev HRFormer-B AP50 92.7 #17 of 47 Archive leaderboard report
Pose Estimation COCO test-dev HRFormer-B AP75 83.8 #17 of 47 Archive leaderboard report
Pose Estimation COCO test-dev HRFormer-B APL 82.3 #17 of 47 Archive leaderboard report
Pose Estimation COCO test-dev HRFormer-B APM 72.5 #17 of 47 Archive leaderboard report
Pose Estimation COCO test-dev HRFormer-B AR 81.2 #17 of 47 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 EncodingsAdamAttentionBPEConvolutionDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxStochastic DepthSwin TransformerTransformerVision Transformer

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