AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

HUMUS-Net: Hybrid Unrolled Multi-scale Network Architecture for Accelerated MRI Reconstruction

Zalan Fabian, Berk Tinaz, Mahdi Soltanolkotabi

NeurIPS 2022 · Architectures and efficiency › Attention and transformers · approach: Efficient image transformers · 35 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
replace single-scale Swin with multi-scale MUST feature extractor (ablation)unrolled SwinIR (Un-SS) · Stanford 3DSSIM: 0.9319 -> 0.9357Improved
increase patch size from 1x1 to 2x2 (with 3x embedding dim) (ablation)Un-MS · Stanford 3DSSIM: 0.9357 -> 0.9171Hurt
process low-resolution convolutional features in transformer (HUMUS-Block) instead of full-resolution (ablation)Un-MS · Stanford 3DSSIM: 0.9357 -> 0.9449Improved
unroll SwinIR with sensitivity map estimator, small size (Un-SS) (ablation)SwinIR · Stanford 3DSSIM: 0.9336 -> 0.9319Hurt
replace U-Net denoiser with HUMUS-Block at matched parameters (ablation)unrolled MRI network · Stanford 3DSSIM: 0.9348 -> 0.9378Improved
add adjacent slice reconstruction to E2E-VarNet (ablation)E2E-VarNet · Stanford 3D x8SSIM: 0.9432 -> 0.9457Improved
HUMUS-Net without ASR vs E2E-VarNet with ASR (ablation)HUMUS-Net · Stanford 3D x8SSIM: 0.9457 -> 0.9467Improved
HUMUS-Net replaces E2E-VarNetfastMRI knee x8 test · fastMRI knee testSSIM: 0.8900 -> 0.8936Improved
HUMUS-Net replaces E2E-VarNetStanford 2D validation · Stanford 2DPSNR: 33.9 -> 33.7Hurt
increase number of unrolled cascades (ablation)HUMUS-Net · Stanford 3D x8SSIMImproved
add validation split to training dataHUMUS-Net · fastMRI knee testSSIM: 0.8936 -> 0.8945Improved
scale HUMUS-Net to HUMUS-Net-L (embed dim 66 to 96)HUMUS-Net · fastMRI knee testSSIM: 0.8936 -> 0.8944Improved

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