Papers › TransNeXt: Robust Foveal Visual Perception for Vision Transformers

TransNeXt: Robust Foveal Visual Perception for Vision Transformers

28 Nov 2023CVPR 2024 1arXiv:2311.17132archive 2025-07-28

Dai Shi

Due to the depth degradation effect in residual connections, many efficient Vision Transformers models that rely on stacking layers for information exchange often fail to form sufficient information mixing, leading to unnatural visual perception. To address this issue, in this paper, we propose Aggregated Attention, a biomimetic design-based token mixer that simulates biological foveal vision and continuous eye movement while enabling each token on the feature map to have a global perception. Furthermore, we incorporate learnable tokens that interact with conventional queries and keys, which further diversifies the generation of affinity matrices beyond merely relying on the similarity between queries and keys. Our approach does not rely on stacking for information exchange, thus effectively avoiding depth degradation and achieving natural visual perception. Additionally, we propose Convolutional GLU, a channel mixer that bridges the gap between GLU and SE mechanism, which empowers each token to have channel attention based on its nearest neighbor image features, enhancing local modeling capability and model robustness. We combine aggregated attention and convolutional GLU to create a new visual backbone called TransNeXt. Extensive experiments demonstrate that our TransNeXt achieves state-of-the-art performance across multiple model sizes. At a resolution of 224², TransNeXt-Tiny attains an ImageNet accuracy of 84.0%, surpassing ConvNeXt-B with 69% fewer parameters. Our TransNeXt-Base achieves an ImageNet accuracy of 86.2% and an ImageNet-A accuracy of 61.6% at a resolution of 384², a COCO object detection mAP of 57.1, and an ADE20K semantic segmentation mIoU of 54.7.

PaperPDFConference PDFCodeCode 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 5 code samples harvested from 2 repositories linked to this paper: 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified. Where Syntology's instrument failed: 2 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): 3 samples from 1 repository, 3 ran (0 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 1 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

daishiresearch/transnext officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 2 samples, 2 ran (0 instrument-failed) · 0 unverified report
Westlake-AI/openmixup mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to mixup: Beyond Empirical Risk Minimization (arXiv:1710.09412; that paper's own run record: 36 ran (of which 5 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 0 violated, 14 with no contract checked; 22 where Syntology's instrument failed) · 11 unverified), Visualizing the Loss Landscape of Neural Nets (arXiv:1712.09913; that paper's own run record: 3 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 5 unverified), CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features (arXiv:1905.04899; that paper's own run record: 17 ran (of which 6 constructed an object rather than computing a result; 10 with no instrument failure: 2 honoured, 0 violated, 8 with no contract checked; 7 where Syntology's instrument failed) · 7 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
chenller/mmseg-extension mentioned on GitHubpytorchSyntology: 3 samples, 3 ran (3 instrument-failed) · 0 unverified report
gitlab.com/birder/birder pytorchSyntology: not harvested 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.

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

5 samples harvested: 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 0 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.

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

Licence: 0 of the 5 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”. “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.

daishiresearch/transnext official repository 2 samples, 2 ran (0 instrument-failed) · 0 unverified
get_relative_position_cpb classification/transnext.py ran Apache-2.0 (permissive) · f67dfbdd557f5e3f · report fixture: signature query_size:tuple/1/int, key_size:tuple/1/int, pretrain_size:tuple/1/int · returned tuple · level recorded 2026-09-24
is_installed in 7 papers classification/transnext.py ran Apache-2.0 (permissive) · df3275abf9c53939 · report fixture: signature package_name:str/0/str · returned bool · level recorded 2026-09-24
chenller/mmseg-extension community (archive-listed) 3 samples, 3 ran (3 instrument-failed) · 0 unverified
get_relative_position_cpb mmsegext/models/backbones/transnext_cuda.py ran; the contract check could not run (RAISES) · metamorphic tier: deterministic Apache-2.0 (permissive) · af9e1bafd1dd18f3 · report fixture: signature query_size:tuple/1/int, key_size:tuple/1/int, pretrain_size:tuple/1/int, device:device/0/? · returned tuple · level recorded 2026-09-22
get_seqlen_and_mask mmsegext/models/backbones/transnext_native.py ran · our draft was wrong Apache-2.0 (permissive) · e076f3ca82515c34 · report fixture: signature input_resolution:tuple/1/int, window_size:int/0/int, device:device/0/? · returned tuple · level recorded 2026-09-22
get_seqlen_scale mmsegext/models/backbones/transnext_cuda.py ran · our draft was wrong Apache-2.0 (permissive) · 973ff8b81a13f50b · report fixture: signature input_resolution:tuple/1/int, window_size:int/0/int, device:str/0/str · returned Tensor · level recorded 2026-09-22

Tasks

ClassificationDomain GeneralizationImage ClassificationObject DetectionSemantic Segmentationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Domain Generalization ImageNet-A TransNeXt-Base (IN-1K supervised, 384) Number of params 89.7M #15 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Base (IN-1K supervised, 384) Top-1 accuracy % 61.6 #15 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Small (IN-1K supervised, 384) Number of params 49.7M #16 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Small (IN-1K supervised, 384) Top-1 accuracy % 58.3 #16 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Base (IN-1K supervised, 224) Number of params 89.7M #19 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Base (IN-1K supervised, 224) Top-1 accuracy % 50.6 #19 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Small (IN-1K supervised, 224) Number of params 49.7M #21 of 39 Archive leaderboard report
Domain Generalization ImageNet-A TransNeXt-Small (IN-1K supervised, 224) Top-1 accuracy % 47.1 #21 of 39 Archive leaderboard report
Image Classification ImageNet TransNeXt-Base (IN-1K supervised, 384) GFLOPs 56.3 #169 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Base (IN-1K supervised, 384) Number of params 89.7M #169 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Base (IN-1K supervised, 384) Top 1 Accuracy 86.2% #169 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Small (IN-1K supervised, 384) GFLOPs 32.1 #179 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Small (IN-1K supervised, 384) Number of params 49.7M #179 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Small (IN-1K supervised, 384) Top 1 Accuracy 86.0% #179 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Small (IN-1K supervised, 224) GFLOPs 10.3 #295 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Small (IN-1K supervised, 224) Number of params 49.7M #295 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Small (IN-1K supervised, 224) Top 1 Accuracy 84.7% #295 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Tiny (IN-1K supervised, 224) GFLOPs 5.7 #365 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Tiny (IN-1K supervised, 224) Number of params 28.2M #365 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Tiny (IN-1K supervised, 224) Top 1 Accuracy 84.0% #365 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Micro (IN-1K supervised, 224) GFLOPs 2.7 #528 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Micro (IN-1K supervised, 224) Number of params 12.8M #528 of 1060 Archive leaderboard report
Image Classification ImageNet TransNeXt-Micro (IN-1K supervised, 224) Top 1 Accuracy 82.5% #528 of 1060 Archive leaderboard report
Object Detection COCO minival TransNeXt-Base (IN-1K pretrain, DINO 1x) box AP 57.1 #42 of 220 Archive leaderboard report
Object Detection COCO minival TransNeXt-Small (IN-1K pretrain, DINO 1x) box AP 56.6 #44 of 220 Archive leaderboard report
Object Detection COCO minival TransNeXt-Tiny (IN-1K pretrain, DINO 1x) box AP 55.7 #48 of 220 Archive leaderboard report
Semantic Segmentation ADE20K TransNeXt-Base (IN-1K pretrain, Mask2Former, 512) Params (M) 109 #55 of 235 Archive leaderboard report
Semantic Segmentation ADE20K TransNeXt-Base (IN-1K pretrain, Mask2Former, 512) Validation mIoU 54.7 #55 of 235 Archive leaderboard report
Semantic Segmentation ADE20K TransNeXt-Small (IN-1K pretrain, Mask2Former, 512) Params (M) 69 #68 of 235 Archive leaderboard report
Semantic Segmentation ADE20K TransNeXt-Small (IN-1K pretrain, Mask2Former, 512) Validation mIoU 54.1 #68 of 235 Archive leaderboard report
Semantic Segmentation ADE20K TransNeXt-Tiny (IN-1K pretrain, Mask2Former, 512) Params (M) 47.5 #82 of 235 Archive leaderboard report
Semantic Segmentation ADE20K TransNeXt-Tiny (IN-1K pretrain, Mask2Former, 512) Validation mIoU 53.4 #82 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.

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