AI Paper Atlas

Research map / Image generation / Image restoration

Unsupervised Non-Rigid Image Distortion Removal via Grid Deformation

Nianyi Li, Simron Thapa, Cameron Whyte, Albert W. Reed, Suren Jayasuriya, Jinwei Ye

ICCV 2021 · Image generation › Image restoration · approach: Reflection removal · 41 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 iterative refinement after initialization (ablation)Our init. · Simulated air turbulence (strong)SSIM: 0.667 -> 0.799Improved
add iterative refinement after initialization (ablation)Our init. · Simulated air turbulence (weak)PSNR: 25.20 -> 24.29Hurt
use 4 conv layers per grid deformer instead of 2 (ablation)grid deformer · Simulated air turbulencePSNR: 19.23 -> 20.48Improved
use 4 conv layers instead of 6 (ablation)grid deformer · Simulated air turbulencePSNR: 20.06 -> 20.48Improved
per-frame grid deformers instead of one autoencoder predicting all grids (ablation)grid deformer · Simulated air turbulencePSNR: 16.83 -> 20.48Improved
increase number of input frames from 2 to 10 (ablation)unsupervised network · Simulated air turbulencePSNR: 17.55 -> 20.50Improved
increase number of input frames from 10 to 20 (ablation)unsupervised network · Simulated air turbulencePSNR: 20.50 -> 21.13No clear effect
increase number of input frames from 10 to 20 (running time) (ablation)unsupervised network · Simulated air turbulencerunning time: 265s -> 499sHurt
add Gaussian random Fourier feature position encoding to image generator (ablation)image generator · Simulated air turbulenceSSIM: 30.9% improvement in SSIM and 13.9% improvement in PSNRImproved
add initialization step for grid deformer and image generator (ablation)unsupervised network · Simulated air turbulencerestoration qualityImproved
add initialization losses to the main optimization loop instead of separate step (ablation)unsupervised network · Simulated air turbulencerestoration qualityHurt
add regularizer R(I_k) (ablation)unsupervised network · Simulated air turbulencePSNR and SSIMImproved

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