AI Paper Atlas

Research map / Point clouds / Point cloud classification

Zero-Shot Learning on 3D Point Cloud Objects and Beyond

Ali Cheraghian, Shafin Rahman, Townim Faisal Chowdhury, Dylan Campbell, Lars Petersson

IJCV 2022 · Point clouds › Point cloud classification · approach: Self-supervised point cloud pretraining · 70 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
project semantics to point-cloud features (S2F) instead of F2S (ablation)inductive PointConv ZSL model · ModelNet10 ZSLaccuracy: 24.45 -> 32.49Improved
add unsupervised triplet loss (transductive) (ablation)S2F model · ModelNet10 ZSLaccuracy: 32.49 -> 43.17Improved
add hubness loss (ablation)transductive model · ModelNet10 ZSLaccuracy: 43.17 -> 61.45Improved
add unbiasing loss (ablation)transductive model · ModelNet10 GZSLharmonic mean: 57.33 -> 73.39Improved
add unbiasing loss (ablation)transductive model · ModelNet10 ZSLaccuracy: 61.45 -> 68.50Improved
add unbiasing loss (ablation)transductive model · ScanObjectNN ZSLaccuracy: 30.74 -> 30.53No clear effect
use confidence weighting pi in hubness loss instead of pi=1 (ablation)transductive model · ModelNet10 ZSLaccuracy: 46.37 -> 68.50Improved
use confidence weighting pi in hubness loss instead of pi=1 (ablation)transductive model · ScanObjectNN GZSLharmonic mean: 28.19 -> 45.63Improved
add transductive training with unlabeled unseen dataOurs · ModelNet10 GZSLharmonic mean: 12.69 -> 73.39Improved
use PointConv backbone instead of PointNetOurs (Transductive) · ModelNet10 ZSLaccuracy: 18.28 -> 68.50Improved
use PointConv backbone instead of DGCNNOurs (Transductive) · ModelNet10 ZSLaccuracy: 60.05 -> 68.50Improved
add transductive learning with few seen classes (20 vs 30)Ours (Transductive) · ModelNet10 ZSLaccuracy: 68.50 -> 27.42Hurt

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