Research map / Point clouds / Point cloud classification
Point2Node: Correlation Learning of Dynamic-Node for Point Cloud Feature Modeling
Wenkai Han, Chenglu Wen, Cheng Wang, Xin Li, Qing Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use Point2Node instead of PointWeb | Point cloud semantic segmentation · S3DIS 6-fold cross-validation | mIoU: 66.73 -> 70.00 | Improved |
| Use Point2Node instead of GACNet | Point cloud semantic segmentation · S3DIS Area 5 | mIoU: 62.85 -> 62.96 | Improved |
| Use Point2Node instead of PointWeb | Semantic voxel labeling · ScanNet | OA: 85.9 -> 86.3 | Improved |
| Use Point2Node instead of A-CNN | Shape classification · ModelNet40, 1k xyz | OA: 92.6 -> 93.0 | Improved |
| Add the self-correlation block (ablation) | χ-conv Encoder-Decoder baseline · S3DIS Area 5 | mIoU: 56.70 -> 57.45 | Improved |
| Add the local-correlation block (ablation) | χ-conv Encoder-Decoder baseline · S3DIS Area 5 | mIoU: 56.70 -> 57.91 | Improved |
| Add the non-local-correlation block (ablation) | χ-conv Encoder-Decoder baseline · S3DIS Area 5 | mIoU: 56.70 -> 59.74 | Improved |
| Use sequential, fully updated correlation blocks instead of placing all blocks in parallel (ablation) | Point2Node · S3DIS Area 5 | mIoU: 59.51 -> 62.96 | Improved |
| Use sequential local and non-local blocks rather than parallel local and non-local blocks (ablation) | Point2Node · S3DIS Area 5 | mIoU: 60.57 -> 62.96 | Improved |
| Replace adaptive aggregation with linear aggregation (ablation) | Point2Node · S3DIS Area 5 | mIoU: 62.96 -> 58.17 | Hurt |
| Use parameterized adaptive aggregation instead of parameter-free adaptive aggregation (ablation) | Point2Node · S3DIS Area 5 | mIoU: 60.57 -> 62.96 | Improved |
Later papers that compared with this method
1 later accepted paper reports Point2Node as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Investigate Indistinguishable Points in Semantic Segmentation of 3D Point Cloud (AAAI 2021 · better on 2 of 2 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- PointWeb — PointWeb: Enhancing Local Neighborhood Features for Point Cloud Processing (CVPR 2019) · better on 2 of 2 shared results
- A-CNN — A-CNN: Annularly Convolutional Neural Networks on Point Clouds (CVPR 2019) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling (CVPR 2018 · similarity 0.87)
- Point2Sequence: Learning the Shape Representation of 3D Point Clouds with an Attention-Based Sequence to Sequence Network (AAAI 2019 · similarity 0.86)
- PointCNN: Convolution On X-Transformed Points (NeurIPS 2018 · similarity 0.86)
- Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline (ICML 2021 · similarity 0.86)
- Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Framework (ICLR 2022 · similarity 0.86)
- Feature Interactive Representation for Point Cloud Registration (ICCV 2021 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
