Semantic Segmentation On S3Dis Area5

評価指標

Number of params

mAcc

mIoU

oAcc

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

モデル名	Number of params	mAcc	mIoU	oAcc	Paper Title	Repository
SPoTr	N/A	76.4	70.8	90.7	Self-positioning Point-based Transformer for Point Cloud Understanding
PointVector-XL	-	78.1	72.3	91	PointVector: A Vector Representation In Point Cloud Analysis
DITR	-	-	74.1	-	-	-
KPConv	14.1M	72.8	67.1	-	KPConv: Flexible and Deformable Convolution for Point Clouds
PointMixer	6.5M	77.4	71.4	-	PointMixer: MLP-Mixer for Point Cloud Understanding
TangentConv	N/A	62.2	-	-	Tangent Convolutions for Dense Prediction in 3D
SSP+SPG	290K	68.2	61.7	87.9	Point Cloud Oversegmentation with Graph-Structured Deep Metric Learning
DPC	N/A	-	61.28	-	Dilated Point Convolutions: On the Receptive Field Size of Point Convolutions on 3D Point Clouds
HPEIN	N/A	68.3	61.85	87.18	Hierarchical Point-Edge Interaction Network for Point Cloud Semantic Segmentation	-
SegCloud	N/A	57.4	48.9	-	SEGCloud: Semantic Segmentation of 3D Point Clouds	-
WindowNorm+PointTransformer	N/A	77.9	71.4	91.1	Window Normalization: Enhancing Point Cloud Understanding by Unifying Inconsistent Point Densities
SuperCluster	0.21	-	68.1	-	Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering
Swin3D-L	N/A	80.5	74.5	92.7	Swin3D: A Pretrained Transformer Backbone for 3D Indoor Scene Understanding
PointNet	N/A	-	41.1	-	Point Transformer
SPG(PTv2)	-	79.5	73.3	91.9	Subspace Prototype Guidance for Mitigating Class Imbalance in Point Cloud Semantic Segmentation
Pamba	-	-	73.5	-	Pamba: Enhancing Global Interaction in Point Clouds via State Space Model	-
ConDaFormer	-	78.9	73.5	92.4	ConDaFormer: Disassembled Transformer with Local Structure Enhancement for 3D Point Cloud Understanding
Serialized Piont Mamba	-	-	70.6	-	Serialized Point Mamba: A Serialized Point Cloud Mamba Segmentation Model	-
SCF-Net	N/A	71.8	63.7	87.2	SCF-Net: Learning Spatial Contextual Features for Large-Scale Point Cloud Segmentation
Superpoint Transformer	212K	77.3	68.9	89.5	Efficient 3D Semantic Segmentation with Superpoint Transformer

0 of 60 row(s) selected.