AI Paper Atlas

Research map / Point clouds / Point cloud classification

Benchmarking and Analyzing Point Cloud Classification under Corruptions

Jiawei Ren, Liang Pan, Ziwei Liu

ICML 2022 · Point clouds › Point cloud classification · approach: Adversarial point cloud attacks · 30 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
Use RPC with 3D representation, k-NN, frequency grouping, and self-attentionPoint cloud classification · ModelNet-CmCE: 0.892 -> 0.863Improved
Add WOLFMix, which deforms point clouds before rigid mixingDGCNN · ModelNet-CmCE: 1.000 -> 0.590Improved
Use WOLFMix instead of standalone PointWOLF or RSMixDGCNN · ModelNet-CmCE: 0.814 -> 0.590; 0.745 -> 0.590Improved
Add WOLFMix augmentationGDANet · ModelNet-CmCE: 0.892 -> 0.571Improved
Add WOLFMix augmentationPointNet · ModelNet-CmCE: 1.422 -> 1.180Improved
Add WOLFMix augmentationPointNet · ModelNet40OA: 0.907 -> 0.884Hurt
Use k-NN instead of ball-query for local aggregationPoint cloud classifiers · ModelNet-C, especially Drop-LmCEImproved
Use ball-query instead of k-NN for local aggregationPoint cloud classifiers · ModelNet-C Add-GmCEImproved
Add advanced grouping, such as frequency grouping or curve groupingPoint cloud classifiers · ModelNet-CmCEImproved
Use self-attention rather than conventional feature operatorsPoint cloud classifiers · ModelNet-C, particularly Drop-GmCEImproved
Project 3D points to 2D images for classificationPoint cloud classifiers · ModelNet-CmCEMixed
Pretrain DGCNN with OcCoDGCNN · ModelNet-CmCE: 1.000 -> 1.047Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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