AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

NCDL: A Framework for Deep Learning on non-Cartesian Lattices

Joshua John Horacsek, Usman Alim

NeurIPS 2023 · Probabilistic and scientific modeling › Molecules and proteins · approach: Sparse graphical and tensor models · 0 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 Cartesian convolutions with quincunx downsampling plus quincunx convolutionconvolutional auto-encoder · CelebAPSNR: 17.53 -> 17.92Improved
Replace Cartesian convolutions with quincunx downsampling plus quincunx convolutionconvolutional auto-encoder · CelebAL1: 0.08804 -> 0.08318Improved
Replace Cartesian convolutions with quincunx downsampling plus quincunx convolutionconvolutional auto-encoder · CelebASSIM: 0.6113 -> 0.6206Improved
Replace Cartesian convolutions with quincunx downsampling plus quincunx convolutionconvolutional auto-encoder · CelebALPIPS: 0.4808 -> 0.4765Improved
Replace Cartesian with quincunx lattice in U-NetU-Net · DUTSBCE: 0.19266 -> 0.18706Improved
Replace Cartesian with quincunx lattice in U-NetU-Net · DUTSSSIM: 0.2900 -> 0.3074Improved
Replace Cartesian with quincunx lattice in U-NetU-Net · DUTSL1: 0.07583 -> 0.08268Hurt
Replace Cartesian with quincunx lattice in U-NetU-Net · DUTSL2: 0.04969 -> 0.05107Hurt
Use NCDL hexagonal convolution instead of HexagDLy at small grid sizeshexagonal convolution · GPU runtime, grid 256-2048mean time (ms)Hurt
Use NCDL hexagonal convolution instead of HexagDLy at large grid sizeshexagonal convolution · GPU runtime, grid 256-2048mean time (ms)Improved
Use NCDL non-Cartesian convolution instead of Cartesian Conv2D1-ring filter · GPU runtimemean time (ms)Mixed

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