AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Learning the Geometry of Wave-Based Imaging

Konik Kothari, Maarten de Hoop, Ivan Dokmanić

NeurIPS 2020 · Probabilistic and scientific modeling › Molecules and proteins · approach: Equivariant and disentangled latent representations · 1 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 U-Net with FIONetReverse time continuation · Shapes, Reflectors, MNIST, and CelebA faces; 3000 training samplesnormalized cross-correlation: 0.21 -> 0.81; 0.07 -> 0.63; 0.35 -> 0.87; 0.27 -> 0.83Improved
Replace U-Net with FIONetReverse time continuation · Thick lines, 3000 samplesnormalized cross-correlation: 0.92 -> 0.97Improved
Replace U-Net with FIONetInverse source problem · Thick lines, Reflectors, Random shapes, and MNISTnormalized cross-correlation: 0.60 -> 0.68; 0.38 -> 0.87; 0.53 -> 0.67; 0.43 -> 0.58Improved
Replace U-Net with FIONetReflector imaging · Reflectors, Shuffled Reflectors, and Random Reflectorsnormalized cross-correlation: 0.42 -> 0.74; 0.41 -> 0.72; 0.38 -> 0.55Improved
Replace U-Net with FIONetReflector imaging · Linesnormalized cross-correlation: 0.77 -> 0.75Hurt
Use FIONet rather than U-NetReflector imaging · Linessmoothed Wasserstein-2: 5.69 -> 6.74Hurt
Use FIONet rather than U-NetReverse time continuation · Out-of-distribution datasetsnormalized cross-correlation: Shapes: 0.21 -> 0.81; Reflectors: 0.07 -> 0.63; MNIST: 0.35 -> 0.87; CelebA faces: 0.27 -> 0.83Improved
Remove the learned multipliers (ablation)FIONet · All problemsperformanceNo clear effect
Pretrain the routing network before full FIONet training (ablation)FIONet · Inverse imaging tasksreconstruction qualityImproved

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