AI Paper Atlas

Research map / Point clouds / Point cloud classification

Adaptive Canonicalization with Application to Invariant Anisotropic Geometric Networks

Ya-Wei Eileen Lin, Ron Levie

ICLR 2026 · Point clouds › Point cloud classification · approach: Equivariant point cloud networks · 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
add adaptive canonicalization to NLSF (A-NLSF)NLSF · Grid signal orientationaccuracy: 50.07 -> 99.38Improved
add adaptive canonicalization to NLSFNLSF · TUDataset ENZYMESaccuracy: 65.94 -> 73.01Improved
add adaptive canonicalization to NLSFNLSF · TUDataset MUTAGaccuracy: 84.13 -> 87.94Improved
use adaptive canonicalization instead of OAP canonicalizationGIN · TUDataset PROTEINSaccuracy: 83.41 -> 85.47Improved
add fixed canonicalization (OAP) to GINGIN · TUDataset ENZYMESaccuracy: 45.69 -> 58.40Improved
add fixed canonicalization (OAP) to GINGIN · Grid signal orientationaccuracy: 50.00 -> 50.03No clear effect
add frame averaging to GINGIN · TUDataset PROTEINSaccuracy: 76.02 -> 79.53Improved
use A-NLSF instead of S2GNNS2GNN · TUDataset NCI1accuracy: 75.62 -> 82.01Improved
use A-NLSF instead of GPSGPS · ogbg-molhivAUROC: 0.7880 -> 0.8019Improved
use A-NLSF instead of OAP canonicalizationGatedGCN · ogbg-ppaaccuracy: 0.7745 -> 0.8149Improved
add adaptive canonicalization to DGCNNDGCNN · ModelNet40accuracy: 89.0 -> 91.6Improved
add adaptive canonicalization to PointNetPointNet · ModelNet40accuracy: 75.8 -> 81.1Improved

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