Research map / Point clouds / Point cloud classification
GraphFit: Learning Multi-Scale Graph-Convolutional Representation for Point Cloud Normal Estimation
Keqiang Li, Mingyang Zhao, Huaiyu Wu, Dong-Ming Yan, Zhen Shen, Fei-Yue Wang, Gang Xiong
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 graph convolution with adaptive module instead of DGCNN backbone (ablation) | GraphFit · PCPNet (average) | angle RMSE: 11.08 -> 10.49 | Improved |
| add adaptive attention module (ablation) | GraphFit (single scale) · PCPNet (average) | angle RMSE: 10.74 -> 10.58 | Improved |
| add adaptive module under Striped density (ablation) | GraphFit · PCPNet (average) Striped | angle RMSE: 5.77 -> 5.38 | Improved |
| add multi-scale representation (n=1) (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.91 -> 10.63 | Improved |
| add multi-scale representation (n=3) (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.58 -> 10.49 | Improved |
| add multi-scale representation (n=4) (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.75 -> 10.44 | Improved |
| increase Jet order from 1 to 4 (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.63 -> 10.44 | Improved |
| increase neighborhood size from 256 to 500 (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.49 -> 10.26 | Improved |
| increase neighborhood size from 500 to 700 (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.26 -> 10.34 | No clear effect |
| increase number of graph blocks from 1 to 3 (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.76 -> 10.52 | Improved |
| increase graph blocks from 2 to 3 (ablation) | GraphFit · PCPNet (average) | angle RMSE: 10.58 -> 10.52 | Improved |
| use GraphFit instead of AdaFit | normal estimator · SceneNN | angle RMSE: 22.61 -> 21.33 | Improved |
Later papers that compared with this method
3 later accepted papers report GraphFit as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- OscuFit: Learning to Fit Osculating Implicit Quadrics for Point Clouds (AAAI 2026 · better on 10 of 10 shared results)
- ESNE: Efficient Surface Normal Estimation for LiDAR Point Clouds with Sequential Modeling and Variability Guidance (ECCV 2026 · better on 2 of 2 shared results)
- Rethinking the Approximation Error in 3D Surface Fitting for Point Cloud Normal Estimation (CVPR 2023 · 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 3 of 3 on most shared results.
- AdaFit — AdaFit: Rethinking Learning-Based Normal Estimation on Point Clouds (ICCV 2021) · better on 9 of 9 shared results
- DeepFit — DeepFit: 3D Surface Fitting via Neural Network Weighted Least Squares (ECCV 2020) · better on 2 of 2 shared results
- Nesti-Net — Nesti-Net: Normal Estimation for Unstructured 3D Point Clouds Using Convolutional Neural Networks (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
- Nesti-Net: Normal Estimation for Unstructured 3D Point Clouds Using Convolutional Neural Networks (CVPR 2019 · similarity 0.89)
- Geometry and Learning Co-Supported Normal Estimation for Unstructured Point Cloud (CVPR 2020 · similarity 0.89)
- AdaFit: Rethinking Learning-Based Normal Estimation on Point Clouds (ICCV 2021 · similarity 0.89)
- Learning Graph-Convolutional Representations for Point Cloud Denoising (ECCV 2020 · similarity 0.89)
- Learning Localized Generative Models for 3D Point Clouds via Graph Convolution (ICLR 2019 · similarity 0.88)
- CMG-Net: Robust Normal Estimation for Point Clouds via Chamfer Normal Distance and Multi-Scale Geometry (AAAI 2024 · similarity 0.87)
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.
