mlsanityData to 7 Oct 2026

FatNet: A Feature-attentive Network for 3D Point Cloud Processing

arXiv 2104.03427 ↗First posted 7 Apr 2021cs.CVcited by 10

Not on the map. Why?

Abstract (arXiv)

The application of deep learning to 3D point clouds is challenging due to its lack of order. Inspired by the point embeddings of PointNet and the edge embeddings of DGCNNs, we propose three improvements to the task of point cloud analysis. First, we introduce a novel feature-attentive neural network layer, a FAT layer, that combines both global point-based features and local edge-based features in order to generate better embeddings. Second, we find that applying the same attention mechanism across two different forms of feature map aggregation, max pooling and average pooling, gives better performance than either alone. Third, we observe that residual feature reuse in this setting propagates information more effectively between the layers, and makes the network easier to train. Our architecture achieves state-of-the-art results on the task of point cloud classification, as demonstrated on the ModelNet40 dataset, and an extremely competitive performance on the ShapeNet part segmentation challenge.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 25 of its 31 references that are in the map. First posted on arXiv 7 Apr 2021.

Builds on

25 of 31 references.

References · 25

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  4. 2017Squeeze-and-Excitation Networkscited by 4,763
  5. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  6. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  7. 2017Deep Setscited by 1,394
  8. 2018PointCNN: Convolution On X-Transformed Pointscited by 764
  9. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
  10. 2015Multi-view Convolutional Neural Networks for 3D Shape Recognitioncited by 564
  11. 2019GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyondcited by 331
  12. 2017Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on Graphscited by 307
13 more
  1. 2018SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filterscited by 261
  2. 2016Volumetric and Multi-View CNNs for Object Classification on 3D Datacited by 240
  3. 2018SO-Net: Self-Organizing Network for Point Cloud Analysiscited by 210
  4. 2018SPLATNet: Sparse Lattice Networks for Point Cloud Processingcited by 158
  5. 2018Point Convolutional Neural Networks by Extension Operatorscited by 154
  6. 2017Escape from Cells: Deep Kd-Networks for the Recognition of 3D Point Cloud Modelscited by 142
  7. 2018Concurrent Spatial and Channel Squeeze & Excitation in Fully Convolutional Networkscited by 120
  8. 2017PCPNET: Learning Local Shape Properties from Raw Point Cloudscited by 97
  9. 2018Recurrent Slice Networks for 3D Segmentation of Point Cloudscited by 82
  10. 2016SyncSpecCNN: Synchronized Spectral CNN for 3D Shape Segmentationcited by 70
  11. 2018Exploring Spatial Context for 3D Semantic Segmentation of Point Cloudscited by 33
  12. 20143D ShapeNets: A Deep Representation for Volumetric Shapescited by 27
  13. 20173D Point Cloud Classification and Segmentation using 3D Modified Fisher Vector Representation for Convolutional Neural Networkscited by 8

Built on by

Cited by year

Citing papers per year: 2021: 2; 2022: 3; 2023: 1; 2024: 0; 2025: 4.

10 papers.

  1. 2021PRA-Net: Point Relation-Aware Network for 3D Point Cloud Analysiscited by 58
  2. 2022The Fully Convolutional Transformer for Medical Image Segmentationcited by 31
  3. 20223DCTN: 3D Convolution-Transformer Network for Point Cloud Classificationcited by 25
  4. 2021CpT: Convolutional Point Transformer for 3D Point Cloud Processingcited by 8
  5. 2023GTNet: Graph Transformer Network for 3D Point Cloud Classification and Semantic Segmentationcited by 3
  6. 20223DGTN: 3D Dual-Attention GLocal Transformer Network for Point Cloud Classification and Segmentationcited by 0
  7. 2025Deeply Dual Supervised learning for melanoma recognitioncited by 0
  8. 2025Edge Detection for Organ Boundaries via Top Down Refinement and SubPixel Upsamplingcited by 0
  9. 2025DualResolution Residual Architecture with Artifact Suppression for Melanocytic Lesion Segmentationcited by 0
  10. 2025VesselRW: Weakly Supervised Subcutaneous Vessel Segmentation via Learned Random Walk Propagationcited by 0