Papers › Twins: Revisiting the Design of Spatial Attention in Vision Transformers

Twins: Revisiting the Design of Spatial Attention in Vision Transformers

28 Apr 2021NeurIPS 2021 12arXiv:2104.13840archive 2025-07-28

Xiangxiang Chu, Zhi Tian, Yuqing Wang, Bo Zhang, Haibing Ren, Xiaolin Wei, Huaxia Xia, Chunhua Shen

Very recently, a variety of vision transformer architectures for dense prediction tasks have been proposed and they show that the design of spatial attention is critical to their success in these tasks. In this work, we revisit the design of the spatial attention and demonstrate that a carefully-devised yet simple spatial attention mechanism performs favourably against the state-of-the-art schemes. As a result, we propose two vision transformer architectures, namely, Twins-PCPVT and Twins-SVT. Our proposed architectures are highly-efficient and easy to implement, only involving matrix multiplications that are highly optimized in modern deep learning frameworks. More importantly, the proposed architectures achieve excellent performance on a wide range of visual tasks, including image level classification as well as dense detection and segmentation. The simplicity and strong performance suggest that our proposed architectures may serve as stronger backbones for many vision tasks. Our code is released at https://github.com/Meituan-AutoML/Twins .

PaperPDFConference PDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2104.13840")

Code

Syntology Ran 0 of 2 code samples harvested from 1 repository linked to this paper; 2 have no recorded run.

By repository: official repository: 2 samples from 1 repository, 0 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Meituan-AutoML/Twins officialmentioned in papermentioned on GitHubpytorchApache-2.0 report
ahmedelmahy/myownvit mentioned on GitHubpytorch report
conceptofmind/Twins-SVT-Flax mentioned on GitHubjaxMIT report
rwightman/pytorch-image-models mentioned on GitHubpytorch report
ttt496/vit-pytorch mentioned on GitHubpytorch report
xiaohu2015/pvt_detectron2 mentioned on GitHubpytorchMIT report
PaddlePaddle/PaddleClas paddleApache-2.0 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

2 samples harvested; 0 ran; 0 honoured the contract we drafted; 2 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

2unverified

Licence: 2 of the 2 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 Meituan-AutoML/Twins. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

build_transform Meituan-AutoML/Twins/datasets.py official repository unverified Apache-2.0 recorded; this copy not marked cleared · pointer only · 3bc74137ab36aa79 · report
create_logger Meituan-AutoML/Twins/logger.py official repository unverified Apache-2.0 recorded; this copy not marked cleared · pointer only · 80450600f7f09b0a · report

Tasks

Image ClassificationSemantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ImageNet Twins-SVT-L GFLOPs 15.1 #401 of 1060 Archive leaderboard report
Image Classification ImageNet Twins-SVT-L Number of params 99.2M #401 of 1060 Archive leaderboard report
Image Classification ImageNet Twins-SVT-L Top 1 Accuracy 83.7% #401 of 1060 Archive leaderboard report
Semantic Segmentation ADE20K Twins-SVT-L (UperNet, ImageNet-1k pretrain) Validation mIoU 50.2 #119 of 235 Archive leaderboard report
Semantic Segmentation ADE20K val Twins-SVT-L (UperNet, ImageNet-1k pretrain) mIoU 50.2 #52 of 95 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

Introduced by this paper: Global Sub-Sampled Attention, Locally-Grouped Self-Attention, Spatially Separable Self-Attention, Twins-SVT

AttentionConditional Positional EncodingDense ConnectionsDepthwise ConvolutionGlobal Sub-Sampled AttentionLayer NormalizationLinear LayerLocally-Grouped Self-AttentionMulti-Head AttentionPositional Encoding GeneratorResidual ConnectionSoftmaxSpatially Separable Self-AttentionTwins-PCPVTTwins-SVTVision 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