Few Shot 3D Point Cloud Classification On 3

评估指标

Overall Accuracy

Standard Deviation

评测结果

各个模型在此基准测试上的表现结果

模型名称	Overall Accuracy	Standard Deviation	Paper Title	Repository
PCP-MAE	93.5	3.7	PCP-MAE: Learning to Predict Centers for Point Masked Autoencoders
GPr-Net + Euc (1024)	62.1	1.9	GPr-Net: Geometric Prototypical Network for Point Cloud Few-Shot Learning
Point-BERT	91.0	5.4	Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling
MaskPoint	91.4	4.0	Masked Discrimination for Self-Supervised Learning on Point Clouds
DGCNN	19.85	6.5	Dynamic Graph CNN for Learning on Point Clouds
Point-M2AE	92.3	4.5	Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training
GPr-Net + Hyp (1024)	70.4	1.8	GPr-Net: Geometric Prototypical Network for Point Cloud Few-Shot Learning
IDPT	92.8	-	Instance-aware Dynamic Prompt Tuning for Pre-trained Point Cloud Models
PointNet	46.60	13.5	PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
Point-RAE	93.3	4.0	Regress Before Construct: Regress Autoencoder for Point Cloud Self-supervised Learning
Point-LGMask	92.6	4.3	Point-LGMask: Local and Global Contexts Embedding for Point Cloud Pre-training with Multi-Ratio Masking
ReCon++	94.5	4.1	ShapeLLM: Universal 3D Object Understanding for Embodied Interaction
GPr-Net + Euc (512)	62.3	2.0	GPr-Net: Geometric Prototypical Network for Point Cloud Few-Shot Learning
Point-MAE	92.6	4.1	Masked Autoencoders for Point Cloud Self-supervised Learning
PointCNN	46.60	4.8	PointCNN: Convolution On X-Transformed Points
3D-JEPA	94.3	3.6	3D-JEPA: A Joint Embedding Predictive Architecture for 3D Self-Supervised Representation Learning	-
PointGPT	94.3	3.3	PointGPT: Auto-regressively Generative Pre-training from Point Clouds
OcCo+DGCNN	82.9	1.3	Unsupervised Point Cloud Pre-Training via Occlusion Completion
PointNet++	23.05	7.0	PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
Point-FEMAE	94.0	-	Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoders

0 of 31 row(s) selected.