AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Stripformer: Strip Transformer for Fast Image Deblurring

Fu-Jen Tsai, Yan-Tsung Peng, Yen-Yu Lin, Chung-Chi Tsai, Chia-Wen Lin

ECCV 2022 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers · 324 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 inter-strip attention to intra-strip attention only (ablation)Stripformer · GoPro testPSNR: 32.84 -> 33.00Improved
add intra-strip attention to inter-strip attention only (ablation)Stripformer · GoPro testPSNR: 32.88 -> 33.00Improved
add conditional positional encoding (ablation)Stripformer · GoPro testPSNR: 33.00 -> 33.03Improved
add contrastive loss (ablation)Stripformer · GoPro testPSNR: 33.03 -> 33.08Improved
use strip attention instead of Swin attention (ablation)Stripformer · GoPro testPSNR: 32.39 -> 33.08Improved
use strip attention instead of CCNet attention (ablation)Stripformer · GoPro testPSNR: 32.73 -> 33.08Improved
use strip attention instead of Twins attention (ablation)Stripformer · GoPro testPSNR: 32.89 -> 33.08Improved
use strip attention instead of Swin attention (ablation)Stripformer · HIDE testPSNR: 30.19 -> 31.03Improved
use strip attention instead of CCNet attention (ablation)Stripformer · HIDE testPSNR: 30.61 -> 31.03Improved
use strip attention instead of Twins attention (ablation)Stripformer · HIDE testPSNR: 30.82 -> 31.03Improved
use Stripformer transformer instead of MPRNet CNNdeblurring network · RealBlur-J testPSNR: 31.76 -> 32.48Improved
use Stripformer instead of IPT (no ImageNet pretraining, 20M vs 114M params)deblurring network · GoPro testPSNR: 32.58 -> 33.08Improved

Later papers that compared with this method

1 later accepted paper reports Stripformer as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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