AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation

Wenqiang Zhang, Zilong Huang, Guozhong Luo, Tao Chen, Xinggang Wang, Wenyu Liu, Gang Yu, Chunhua Shen

CVPR 2022 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers · 337 citations (OpenAlex)

Check 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
use multi-scale token pyramid rather than only the last-scale token as Semantics Extractor input (ablation)TopFormer-B · ADE20K valmIoU: 36.6 -> 37.8Improved
inject semantics at scales 1/4, 1/8, 1/16, and 1/32 rather than at 1/8, 1/16, and 1/32 (ablation)TopFormer-B · ADE20K valmIoU: 37.8 -> 38.3Improved
add injection at the 1/8 scale to the 1/16 and 1/32 scales (ablation)TopFormer-B · ADE20K valmIoU: 36.4 -> 37.8Improved
add the Scale-aware Semantics Extractor (ablation)TopFormer-T · ADE20K valmIoU: 22.5 -> 32.8Improved
replace MHSA modules with additional FFNs (ablation)TopFormer-T · ADE20K valmIoU: 32.8 -> 30.4Hurt
use SASE instead of PSP contextual modeling (ablation)TopFormer-T · ADE20K valmIoU: 27.4 -> 32.8Improved
use SASE instead of ASPP contextual modeling (ablation)TopFormer-T · ADE20K valmIoU: 28.2 -> 32.8Improved
combine SigmoidAttn with SemInfo rather than using either alone (ablation)TopFormer · ADE20K valmIoU: 37.3 -> 37.8; 37.0 -> 37.8Improved
use the proposed segmentation head instead of Concat Head (ablation)TopFormer · ADE20K valmIoU: 37.1 -> 37.8Improved
use the proposed segmentation head instead of Sum Head (ablation)TopFormer · ADE20K valmIoU: 37.0 -> 37.8Improved
increase SIM width from 128 to 192 or 256 (ablation)TopFormer · ADE20K valmIoU: 37.3 -> 37.4; 37.3 -> 37.8Improved
increase pooled-token output stride from s128 to s64 (ablation)TopFormer · ADE20K valmIoU: 35.7 -> 37.8Improved

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