AI Paper Atlas

Research map / Image generation / Super-resolution

Rank-One Network: An Effective Framework for Image Restoration

Shangqi Gao, Xiahai Zhuang

TPAMI 2020 · Image generation › Super-resolution · approach: Real-world super-resolution · 9 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
use 3 rank-one projection units instead of 1 (ablation)Realistic image SR, DIV2K RONet · Realistic SR ablationPSNR: 24.31 -> 24.33Improved
use 6 rank-one projection units instead of 3 (ablation)RONet · Realistic SR ablationPSNR: 24.33 -> 24.21Hurt
use L1 loss instead of L2 (ablation)RONet · Realistic SR ablationPSNR: 24.33 -> 24.50Improved
use more residual blocks (12,24,3) or (8,16,8) instead of (3,6,3) (ablation)RONet · Realistic SR ablationPSNR: 24.50 -> 24.53No clear effect
use Xavier uniform instead of MSRA normal initialization (ablation)RONet · Realistic SR ablationPSNR: 24.46 -> 24.50Improved
add batch normalization (ablation)RONet · Realistic SR ablationSSIM: 0.5802 -> 0.5814Improved
multiply residual branch by 0.1 (residual scaling) (ablation)RONet · Realistic SR ablationPSNR: 24.46 -> 24.42Hurt
use unsupervised instead of supervised RO decomposition (ablation)RONet · Realistic SR ablationPSNR: 24.50 -> 24.50No clear effect
use learned RO decomposition instead of SVD (ablation)RONet · Realistic SR ablationtraining time: 99.4h -> 17.0hImproved
train end-to-end instead of pretraining RODec (ablation)RONet · Realistic SR ablationPSNR: 24.50 -> 24.47No clear effect
add deep supervision lambda=0.001 versus lambda=0 (ablation)RONet-R · Realistic SR DIV2K valPSNR: 24.50 -> 24.53Improved
use RONet-R instead of EDSR/WDSRrealistic SR · Realistic SR DIV2K valPSNR: 24.42 -> 24.53Improved

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