AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

A Survey on Vision Transformer

Kai Han, Yunhe Wang, Hanting Chen, Xinghao Chen, Jianyuan Guo, Zhenhua Liu, Yehui Tang, An Xiao, Chunjing Xu, Yixing Xu, Zhaohui Yang, Yiman Zhang and 1 more

TPAMI 2022 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers · 5,163 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add token-based distillation from a CNN teacherDeiT-B · ImageNet-1k valtop-1 accuracy (%): 83.1 -> 84.40Improved
pre-train on JFT-300M instead of training on ImageNet onlyViT · ImageNettop-1 accuracy (%): 88.36Improved
replace convolutions with deformable attention in DETRDeformable DETR · COCO 2017 valAP: 42.0 -> 46.2Improved
add convolution to transformer (CNN + Transformer hybrids)vision backbones · ImageNet-1k valtop-1 accuracy (%): 83.3 -> 84.5Improved
use pyramid architecture with local windows (Swin-B) vs plain DeiT-BSwin-B · ImageNet-1k valtop-1 accuracy (%): 81.8 -> 83.3Improved
scale VOLO from D1 to D5VOLO · ImageNet-1k valtop-1 accuracy (%): 84.2 -> 86.1Improved
train ViT on mid-sized ImageNet instead of large-scale dataViT · ImageNettop-1 accuracy (%)Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy