AI Paper Atlas

Research map / Image generation / Image restoration

Towards Architecture-Agnostic Untrained Networks Priors for Image Reconstruction with Frequency Regularization

Yilin Liu, Yunkui Pang, Jiang Li, Yong Chen, Pew-Thian Yap

ECCV 2024 · Image generation › Image restoration · approach: Inverse imaging with learned priors

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
replace white-noise input with Gaussian-blurred noise input (ablation)A2_256 untrained network · fastMRI brain 4xPSNR: 29.08 -> 33.34Improved
replace noise input with Fourier features of narrower bandwidth (L=4 vs L=16) (ablation)A2_256 · fastMRI brain 4xPSNR: 30.57 -> 31.92Improved
add Kaiser-based bandwidth-controllable upsampler (ablation)deeper architectures (A8_64) · fastMRI brain 4xPSNR: 33.14 -> 33.85 (A8_64); 31.70 -> 33.09 (A8_256)Improved
add Kaiser upsampler on top of blur + Lipschitz for shallow wide network (ablation)A2_256 · fastMRI brain 4xPSNR: 32.90 -> 32.50No clear effect
add Lipschitz regularization alone (ablation)untrained network · fastMRI brain 4xPSNR: 29.41 -> 29.73 (A2_64); 29.08 -> 30.92 (A2_256)Mixed
add TV regularization (ablation)untrained network · fastMRI brain 4xPSNR: 29.41 -> 29.61 (A2_64)No clear effect
apply proposed frequency regularization to compact A2_64A2_64 · fastMRI knee 4xPSNR: 27.62 -> 32.07Improved
add self-validation early stopping to the regularized network (ablation)A2_64 (Ours) · fastMRI knee / Stanford FSEruntime / PSNR: runtime 6.4 -> 0.3 mins; PSNR 32.07 -> 31.97Improved
apply early stopping to unregularized A2_64 (ablation)A2_64 · fastMRI kneePSNR: 27.62 -> 29.59, still below 32.07Mixed
increase upsampler attenuation from none to bilinear/Lanczos-style (ablation)Deep Decoder · fastMRI brain 4xPSNR: 24.55 -> 32.68Improved

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