mlsanityData to 7 Oct 2026

Diffusion Unit: Interpretable Edge Enhancement and Suppression Learning for 3D Point Cloud Segmentation

arXiv 2209.09483 ↗First posted 20 Sep 2022cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

3D point clouds are discrete samples of continuous surfaces which can be used for various applications. However, the lack of true connectivity information, i.e., edge information, makes point cloud recognition challenging. Recent edge-aware methods incorporate edge modeling into network designs to better describe local structures. Although these methods show that incorporating edge information is beneficial, how edge information helps remains unclear, making it difficult for users to analyze its usefulness. To shed light on this issue, in this study, we propose a new algorithm called Diffusion Unit (DU) that handles edge information in a principled and interpretable manner while providing decent improvement. First, we theoretically show that DU learns to perform task-beneficial edge enhancement and suppression. Second, we experimentally observe and verify the edge enhancement and suppression behavior. Third, we empirically demonstrate that this behavior contributes to performance improvement. Extensive experiments and analyses performed on challenging benchmarks verify the effectiveness of DU. Specifically, our method achieves state-of-the-art performance in object part segmentation using ShapeNet part and scene segmentation using S3DIS. Our source code is available at https://github.com/martianxiu/DiffusionUnit.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 50 of its 79 references that are in the map. First posted on arXiv 20 Sep 2022.

Builds on

50 of 79 references.

References · 50

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  5. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  6. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  7. 2017Squeeze-and-Excitation Networkscited by 4,763
  8. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  9. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  10. 2020Point Transformercited by 1,462
  11. 2021MLP-Mixer: An all-MLP Architecture for Visioncited by 1,441
  12. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
38 more
  1. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  2. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  3. 20173D Semantic Segmentation with Submanifold Sparse Convolutional Networkscited by 746
  4. 2021Diffusion Probabilistic Models for 3D Point Cloud Generationcited by 734
  5. 2018PointConv: Deep Convolutional Networks on 3D Point Cloudscited by 656
  6. 2021Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modelingcited by 647
  7. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
  8. 2015Multi-view Convolutional Neural Networks for 3D Shape Recognitioncited by 564
  9. 2019DeepGCNs: Can GCNs Go as Deep as CNNs?cited by 527
  10. 20213D Shape Generation and Completion through Point-Voxel Diffusioncited by 516
  11. 2022PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategiescited by 432
  12. 2019Diffusion Improves Graph Learningcited by 414
  13. 2019Relation-Shape Convolutional Neural Network for Point Cloud Analysiscited by 346
  14. 2015Diffusion-Convolutional Neural Networkscited by 268
  15. 2021GRAND: Graph Neural Diffusioncited by 258
  16. 2020PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Samplingcited by 238
  17. 2021Group-Free 3D Object Detection via Transformerscited by 215
  18. 2021PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Cloudscited by 211
  19. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  20. 2017Octree Generating Networks: Efficient Convolutional Architectures for High-resolution 3D Outputscited by 189
  21. 2017SEGCloud: Semantic Segmentation of 3D Point Cloudscited by 178
  22. 2021Walk in the Cloud: Learning Curves for Point Clouds Shape Analysiscited by 172
  23. 2022Stratified Transformer for 3D Point Cloud Segmentationcited by 166
  24. 2019Modeling Point Clouds with Self-Attention and Gumbel Subset Samplingcited by 151
  25. 2017Escape from Cells: Deep Kd-Networks for the Recognition of 3D Point Cloud Modelscited by 142
  26. 2020Deep Learning for Image and Point Cloud Fusion in Autonomous Driving: A Reviewcited by 107
  27. 2019ShellNet: Efficient Point Cloud Convolutional Neural Networks using Concentric Shells Statisticscited by 104
  28. 2021A Conditional Point Diffusion-Refinement Paradigm for 3D Point Cloud Completioncited by 100
  29. 2020A Closer Look at Local Aggregation Operators in Point Cloud Analysiscited by 87
  30. 2019Interpolated Convolutional Networks for 3D Point Cloud Understandingcited by 86
  31. 2022Surface Representation for Point Cloudscited by 84
  32. 2021PointMixer: MLP-Mixer for Point Cloud Understandingcited by 71
  33. 2019Hierarchical Point-Edge Interaction Network for Point Cloud Semantic Segmentationcited by 68
  34. 2020JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Cloudscited by 62
  35. 2022Contrastive Boundary Learning for Point Cloud Segmentationcited by 56
  36. 2014Rigid-Motion Scattering for Texture Classificationcited by 52
  37. 2018Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNNcited by 33
  38. 2022Enhancing Local Feature Learning for 3D Point Cloud Processing using Unary-Pairwise Attentioncited by 3

Built on by

Not cited yet.