Sonata: Self-Supervised Learning of Reliable Point Representations | ✓ Link | 82.3 | 89.9 | 93.3 | | 128M | | | Sonata + PTv3 | 2025-03-20 |
Point Transformer V3: Simpler, Faster, Stronger | ✓ Link | 80.8 | 87.7 | 92.6 | | 24.1M | | | PTv3 + PPT | 2023-12-15 |
PonderV2: Pave the Way for 3D Foundation Model with A Universal Pre-training Paradigm | ✓ Link | 79.9 | 86.5 | 92.5 | | | | | PonderV2 + SparseUNet | 2023-10-12 |
Swin3D: A Pretrained Transformer Backbone for 3D Indoor Scene Understanding | ✓ Link | 79.8 | 88.0 | 92.4 | | N/A | | | Swin3D-L | 2023-04-14 |
PointVector: A Vector Representation In Point Cloud Analysis | ✓ Link | 78.4 | 86.1 | 91.9 | | | | 24.1 | PointVector-XL | 2022-05-21 |
Towards Large-scale 3D Representation Learning with Multi-dataset Point Prompt Training | ✓ Link | 78.1 | 85.4 | 92.2 | | N/A | | | PPT + SparseUNet | 2023-08-18 |
Window Normalization: Enhancing Point Cloud Understanding by Unifying Inconsistent Point Densities | ✓ Link | 77.6 | 85.8 | 91.7 | | 8.2M | | 8.2 | WindowNorm+StratifiedTransformer | 2022-12-05 |
A Unified Query-based Paradigm for Point Cloud Understanding | ✓ Link | 77.5 | | | | N/A | | | EQ-Net | 2022-03-02 |
Meta Architecture for Point Cloud Analysis | ✓ Link | 77.0 | | 91.3 | 11.0G | 19.7M | | 19.7 | PointMetaBase-XXL | 2022-11-26 |
Efficient 3D Semantic Segmentation with Superpoint Transformer | ✓ Link | 76.0 | 85.8 | 90.4 | | 0.212M | 76.0 | 0.212 | Superpoint Transformer | 2023-06-13 |
Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering | ✓ Link | 75.3 | | | | 0.21M | | 0.21 | SuperCluster | 2024-01-12 |
PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies | ✓ Link | 74.9 | 83.0 | 90.3 | 84.8G | 41.6M | | 41.6 | PointNeXt-XL | 2022-06-09 |
Learning Multi-View Aggregation In the Wild for Large-Scale 3D Semantic Segmentation | ✓ Link | 74.7 | 83.8 | 90.1 | | 41.2M | | 41.2 | DeepViewAgg | 2022-04-15 |
Surface Representation for Point Clouds | ✓ Link | 74.3 | 82.6 | 90.8 | 1.04G | 0.97M | | 0.97 | RepSurf-U | 2022-05-11 |
Window Normalization: Enhancing Point Cloud Understanding by Unifying Inconsistent Point Densities | ✓ Link | 74.1 | 82.5 | 90.2 | | 8.0M | | 8 | WindowNorm+PointTransformer | 2022-12-05 |
PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies | ✓ Link | 73.9 | 82.2 | 89.9 | 15.2G | 7.1M | | 7.1 | PointNeXt-L | 2022-06-09 |
Point Transformer | ✓ Link | 73.5 | 81.9 | 90.2 | | 7.8M | | 7.8 | PointTransformer | 2020-12-16 |
Contrastive Boundary Learning for Point Cloud Segmentation | ✓ Link | 73.1 | 79.4 | 89.6 | | N/A | | | CBL | 2022-03-10 |
Semantic Segmentation for Real Point Cloud Scenes via Bilateral Augmentation and Adaptive Fusion | ✓ Link | 72.2 | 83.1 | 88.9 | | N/A | | | BAAF-Net | 2021-03-12 |
SCF-Net: Learning Spatial Contextual Features for Large-Scale Point Cloud Segmentation | ✓ Link | 71.6 | 82.7 | 88.4 | | N/A | | | SCF-Net | 2021-06-19 |
FG-Net: Fast Large-Scale LiDAR Point Clouds Understanding Network Leveraging Correlated Feature Mining and Geometric-Aware Modelling | ✓ Link | 70.8 | 82.9 | 88.2 | | N/A | | | Feature Geometric Net (FG-Net) | 2020-12-17 |
Learning Inner-Group Relations on Point Clouds | ✓ Link | 70.8 | | | | N/A | | | RPNet | 2021-08-27 |
KPConv: Flexible and Deformable Convolution for Point Clouds | ✓ Link | 70.6 | 79.1 | | | 14.1M | | 14.1 | KPConv | 2019-04-18 |
Point Transformer | ✓ Link | 70.6 | | | | 14.1M | | 14.1 | KPConv | 2020-12-16 |
Fast Point Transformer | ✓ Link | 70.3 | | | | N/A | | | FastPointTrans. (small) | 2021-12-09 |
MuGNet: Multi-Resolution Graph Neural Network for Large-Scale Pointcloud Segmentation | ✓ Link | 69.8 | | 88.5 | | N/A | | | MuGNet | 2020-11-16 |
PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling | ✓ Link | 68.7 | 79.0 | 88.8 | | N/A | | | PointASNL | 2020-03-01 |
Point Cloud Oversegmentation with Graph-Structured Deep Metric Learning | ✓ Link | 68.4 | 78.3 | 87.9 | | 0.290M | | 0.29 | SSP+SPG | 2019-04-03 |
FKAConv: Feature-Kernel Alignment for Point Cloud Convolution | ✓ Link | 68.4 | | | | N/A | | | FKAConv | 2020-04-09 |
ConvPoint: Continuous Convolutions for Point Cloud Processing | ✓ Link | 68.2 | | 88.8 | | 4.7M | | 4.1 | ConvPoint | 2019-04-04 |
JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds | ✓ Link | 67.7 | | | | N/A | | | JSENet | 2020-07-14 |
Cloud Transformers: A Universal Approach To Point Cloud Processing Tasks | | 67.4 | | | | N/A | | | CT2 | 2020-07-22 |
ShellNet: Efficient Point Cloud Convolutional Neural Networks using Concentric Shells Statistics | ✓ Link | 66.8 | | | | N/A | | | ShellNet | 2019-08-17 |
PointWeb: Enhancing Local Neighborhood Features for Point Cloud Processing | ✓ Link | 66.7 | 76.2 | 87.3 | | N/A | | | PointWeb | 2019-06-01 |
A-CNN: Annularly Convolutional Neural Networks on Point Clouds | ✓ Link | 65.4 | | 88.1 | | N/A | | | PointCNN | 2019-04-16 |
Point Transformer | ✓ Link | 65.4 | | | | N/A | | | PointCNN | 2020-12-16 |
4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks | ✓ Link | 65.4 | | | | 37.9M | | 37.9 | MinkowskiNet | 2019-04-18 |
DSPoint: Dual-scale Point Cloud Recognition with High-frequency Fusion | ✓ Link | 63.3 | 70.9 | | | N/A | | | DSPoint | 2021-11-19 |
A-CNN: Annularly Convolutional Neural Networks on Point Clouds | ✓ Link | 62.9 | | 87.3 | | N/A | | | A-CNN | 2019-04-16 |
Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs | ✓ Link | 62.1 | 73 | 85.5 | | 0.290M | | 0.29 | SPG | 2017-11-27 |
A-CNN: Annularly Convolutional Neural Networks on Point Clouds | ✓ Link | 62.1 | | 85.5 | | N/A | | | SPGraph | 2019-04-16 |
Point Transformer | ✓ Link | 62.1 | | | | N/A | | | SPGraph | 2020-12-16 |
JSNet: Joint Instance and Semantic Segmentation of 3D Point Clouds | ✓ Link | 61.7 | 71.7 | 88.7 | | N/A | | | JSNet | 2019-12-20 |
DeepGCNs: Making GCNs Go as Deep as CNNs | ✓ Link | 60.0 | | 85.9 | | N/A | | | DeepGCN | 2019-10-15 |
Associatively Segmenting Instances and Semantics in Point Clouds | ✓ Link | 59.3 | | | | N/A | | | ASIS | 2019-02-26 |
Recurrent Slice Networks for 3D Segmentation of Point Clouds | ✓ Link | 56.5 | 66.5 | | | N/A | | | RSNet | 2018-02-13 |
A-CNN: Annularly Convolutional Neural Networks on Point Clouds | ✓ Link | 56.3 | | 86.9 | | N/A | | | 3P-RNN | 2019-04-16 |
Point-PlaneNet: Plane kernel based convolutional neural network for point clouds analysis | ✓ Link | 54.8 | | 83.9 | | N/A | | | Point-PlaneNet | 2020-05-01 |
Self-supervised Point Cloud Representation Learning via Separating Mixed Shapes | ✓ Link | 51.74 | | | | N/A | | | SMS | 2021-09-01 |
A-CNN: Annularly Convolutional Neural Networks on Point Clouds | ✓ Link | 47.6 | | 78.5 | | N/A | | | PointNet | 2019-04-16 |
Point Transformer | ✓ Link | 47.6 | | | | N/A | | | PointNet | 2020-12-16 |
RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds | ✓ Link | | 81.5 | 87.1 | | 1.2M | 68.5 | 1.2 | RandLA-Net | 2019-11-25 |
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation | ✓ Link | | 66.2 | | | N/A | | | PointNet | 2016-12-02 |
Fully Automated Scan-to-BIM Via Point Cloud Instance Segmentation | ✓ Link | | | | | | 59.5 | | BIM-Net | 2023-09-11 |