List your company
ResearchFormerly WizWand

Few-shot 3D Object Classification (10-way) on ModelNet40 (test)

93.6Accuracy (10-shot)RECON
15.54835.811556.07576.3385Feb 5, 2023May 9, 2023Aug 11, 2023Nov 13, 2023Feb 15, 2024May 19, 2024Aug 21, 2024
Updated 8mo ago

Evaluation results

MethodLinks
RECONTransfer Protocol=MLP-LINEAR, Backbone=RECON-block, Voting Strategy=False2023.0293.695.4
RECONTransfer Protocol=MLP-3, Backbone=RECON-block, Voting Strategy=False2023.0293.695.7
ACTReference=ICLR 23, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.0893.395.6
RECONReference=ICML 23, Tuning Paradigm=Self-Supervised Representation Learning (FULL), Context=baseline2024.0893.395.8
ACTTransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.0293.395.6
RECONTransfer Protocol=FULL, Backbone=RECON-block, Voting Strategy=False2023.0293.395.8
VPPReference=NeurIPS 23, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.089395.4
RECON w/ IDPTReference=ICCV 23, Tuning Paradigm=Parameter-Efficient Supervised Fine-tuning, reproduced=true2024.0892.895.5
Point-MAEReference=ECCV 22, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.0892.695
I2P-MAEReference=CVPR 23, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.0892.695.5
Point-MAETransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.0292.695
Point-MAE†Transfer Protocol=FULL, Backbone=RECON-block, Voting Strategy=False2023.0292.595.2
ACTTransfer Protocol=MLP-3, Backbone=Transformer, Voting Strategy=False2023.0292.494.7
Point-M2AEReference=NeurIPS 22, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.0892.395
Point-M2AETransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.0292.395
RECON w/ PPTReference=ACMMM 25, Tuning Paradigm=Parameter-Efficient Supervised Fine-tuning2024.0892.295.6
RECON w/ DAPTReference=CVPR 24, Tuning Paradigm=Parameter-Efficient Supervised Fine-tuning, reproduced=true2024.0891.994.6
MaskPointTransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.0291.493.4
Point-BERTReference=CVPR 22, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.089192.7
Point-BERTTransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.029192.7
Point-MAETransfer Protocol=MLP-3, Backbone=Transformer, Voting Strategy=False2023.0290.693.8
OcCoReference=ICCV 21, Tuning Paradigm=Self-Supervised Representation Learning (FULL)2024.0889.492.4
OcCoTransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.0289.492.4
Transformer†Transfer Protocol=FULL, Backbone=RECON-block, Voting Strategy=False2023.0285.289.9
TransformerTransfer Protocol=FULL, Backbone=Transformer, Voting Strategy=False2023.0284.689.4
ACTTransfer Protocol=MLP-LINEAR, Backbone=Transformer, Voting Strategy=False2023.0284.590.7
Point-MAETransfer Protocol=MLP-LINEAR, Backbone=Transformer, Voting Strategy=False2023.0283.589.7
OcCoReference=ICCV 21, Tuning Paradigm=Supervised2024.0882.986.5
OcCoTransfer Protocol=None, Backbone=Transformer, Voting Strategy=False2023.0282.986.5
GPr-Net (Hyp)Num Points=512, #Params=1.24K, F/B pass=50KB, Distance Metric=Hyperbolic2023.0471.5973.78
GPr-Net (Hyp)Num Points=1024, #Params=1.24K, F/B pass=50KB, Distance Metric=Hyperbolic2023.0470.4272.83
Enrich-FeaturesNum Points=10242023.0468.7680.72
GPr-Net (Euc)Num Points=512, #Params=1.24K, F/B pass=50KB, Distance Metric=Euclidean2023.0462.3163.33
GPr-Net (Euc)Num Points=1024, #Params=1.24K, F/B pass=50KB, Distance Metric=Euclidean2023.0462.1463.43
SS-FSL (PointNet)Num Points=1024, #Params=3.47M, F/B pass=8.5GB2023.0449.1550.1
SS-FSL (DGCNN)Num Points=1024, #Params=1.82M, F/B pass=53GB2023.0448.553
PointCNNNum Points=10242023.0446.649.95
PointNetNum Points=1024, #Params=3.47M, F/B pass=8.5GB2023.0446.635.2
3D-GANNum Points=10242023.0440.2548.35
PointCapsNetNum Points=1024, #Params=2.15M, F/B pass=39GB2023.043827.15
Latent-GANNum Points=10242023.0432.925.45
PointNet++Num Points=1024, #Params=1.48M, F/B pass=149GB2023.0423.0518.8
DGCNNReference=TOG 19, Tuning Paradigm=Supervised2024.0819.916.9
DGCNNTransfer Protocol=None, Backbone=DGCNN, Voting Strategy=False2023.0219.916.9
DGCNNNum Points=1024, #Params=1.82M, F/B pass=53GB2023.0419.8516.85
FoldingNetNum Points=1024, #Params=0.67M, F/B pass=5.7GB2023.0418.5515.44