List your company
ResearchFormerly WizWand

Object Recognition on ModelNet40 fully-supervised (test)

95.2Overall Accuracy (OA)Point-BERT + MCFT
88.9690.5892.293.82Mar 24, 2026
Updated 7mo ago

Evaluation results

MethodLinks
Point-BERT + MCFTBackbone=Point-BERT, #TP (M)=22.7, FLOPs=4.82026.0395.2
RECON + MCFTBackbone=RECON, #TP (M)=22.7, FLOPs=4.82026.0395.2
Point-MAE + MCFTBackbone=Point-MAE, #TP (M)=22.7, FLOPs=4.82026.0394.9
RECON + MoSTBackbone=RECON, #TP (M)=23.4, FLOPs=4.82026.0394.7
PointMLP#TP (M)=13.2, FLOPs=31.42026.0394.5
Point-MAE + IDPTBackbone=Point-MAE, #TP (M)=23.8, FLOPs=7.22026.0394.4
RECON + DAPTBackbone=RECON, #TP (M)=23.2, FLOPs=5.02026.0394.1
RECON + PointGSTBackbone=RECON, #TP (M)=22.7, FLOPs=4.82026.0394.1
Point-M2AE#TP (M)=15.3, FLOPs=3.62026.0394
Point-MAE + DAPTBackbone=Point-MAE, #TP (M)=23.2, FLOPs=5.02026.0394
Point-MAE + PointGSTBackbone=Point-MAE, #TP (M)=22.7, FLOPs=4.82026.0394
RECON#TP (M)=43.6, FLOPs=5.32026.0393.9
MaskPoint#TP (M)=22.12026.0393.8
Point-MAE#TP (M)=22.1, FLOPs=4.82026.0393.8
Point-BERT + PointGSTBackbone=Point-BERT, #TP (M)=22.7, FLOPs=4.82026.0393.8
ACT#TP (M)=22.1, FLOPs=4.82026.0393.7
Point-BERT + DAPTBackbone=Point-BERT, #TP (M)=23.2, FLOPs=5.02026.0393.6
RECON + IDPTBackbone=RECON, #TP (M)=23.8, FLOPs=7.22026.0393.5
Point-BERT + IDPTBackbone=Point-BERT, #TP (M)=23.8, FLOPs=7.22026.0393.4
Point-MAE + PPTBackbone=Point-MAE, #TP (M)=23.4, FLOPs=4.82026.0393.4
RECON + PPTBackbone=RECON, #TP (M)=23.22026.0393.4
Point-BERT#TP (M)=22.1, FLOPs=4.82026.0393.2
DGCNN#TP (M)=1.8, FLOPs=2.42026.0392.9
OcCo#TP (M)=22.1, FLOPs=4.82026.0392.1
PointNet++#TP (M)=1.5, FLOPs=1.72026.0390.7
PointNet#TP (M)=3.5, FLOPs=0.52026.0389.2