mlsanityData to 7 Oct 2026

MKConv: Multidimensional Feature Representation for Point Cloud Analysis

arXiv 2107.12655 ↗First posted 27 Jul 2021cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

Despite the remarkable success of deep learning, an optimal convolution operation on point clouds remains elusive owing to their irregular data structure. Existing methods mainly focus on designing an effective continuous kernel function that can handle an arbitrary point in continuous space. Various approaches exhibiting high performance have been proposed, but we observe that the standard pointwise feature is represented by 1D channels and can become more informative when its representation involves additional spatial feature dimensions. In this paper, we present Multidimensional Kernel Convolution (MKConv), a novel convolution operator that learns to transform the point feature representation from a vector to a multidimensional matrix. Unlike standard point convolution, MKConv proceeds via two steps. (i) It first activates the spatial dimensions of local feature representation by exploiting multidimensional kernel weights. These spatially expanded features can represent their embedded information through spatial correlation as well as channel correlation in feature space, carrying more detailed local structure information. (ii) Then, discrete convolutions are applied to the multidimensional features which can be regarded as a grid-structured matrix. In this way, we can utilize the discrete convolutions for point cloud data without voxelization that suffers from information loss. Furthermore, we propose a spatial attention module, Multidimensional Local Attention (MLA), to provide comprehensive structure awareness within the local point set by reweighting the spatial feature dimensions. We demonstrate that MKConv has excellent applicability to point cloud processing tasks including object classification, object part segmentation, and scene semantic segmentation with superior results.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 26 of its 42 references that are in the map. First posted on arXiv 27 Jul 2021.

Builds on

26 of 42 references.

References · 26

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  4. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  5. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  6. 2020Point Transformercited by 1,462
  7. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  8. 20173D Semantic Segmentation with Submanifold Sparse Convolutional Networkscited by 746
  9. 2020PCT: Point cloud transformercited by 730
  10. 2018PointConv: Deep Convolutional Networks on 3D Point Cloudscited by 656
  11. 2019Relation-Shape Convolutional Neural Network for Point Cloud Analysiscited by 346
  12. 2018SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filterscited by 261
14 more
  1. 2020PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Samplingcited by 238
  2. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  3. 2021PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Cloudscited by 211
  4. 2018SO-Net: Self-Organizing Network for Point Cloud Analysiscited by 210
  5. 2017SEGCloud: Semantic Segmentation of 3D Point Cloudscited by 178
  6. 2018SPLATNet: Sparse Lattice Networks for Point Cloud Processingcited by 158
  7. 2017Escape from Cells: Deep Kd-Networks for the Recognition of 3D Point Cloud Modelscited by 142
  8. 2019ShellNet: Efficient Point Cloud Convolutional Neural Networks using Concentric Shells Statisticscited by 104
  9. 2019DensePoint: Learning Densely Contextual Representation for Efficient Point Cloud Processingcited by 104
  10. 2019Interpolated Convolutional Networks for 3D Point Cloud Understandingcited by 86
  11. 2018Monte Carlo Convolution for Learning on Non-Uniformly Sampled Point Cloudscited by 62
  12. 2019ConvPoint: Continuous Convolutions for Point Cloud Processingcited by 62
  13. 2018PointSIFT: A SIFT-like Network Module for 3D Point Cloud Semantic Segmentationcited by 58
  14. 2020FPConv: Learning Local Flattening for Point Convolutioncited by 55

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 1; 2023: 4.

6 papers.

  1. 2023Point Cloud Pre-training with Diffusion Modelscited by 47
  2. 2023Advancements in Point Cloud Data Augmentation for Deep Learning: A Surveycited by 24
  3. 2023Joint Representation Learning for Text and 3D Point Cloudcited by 14
  4. 2021LATFormer: Locality-Aware Point-View Fusion Transformer for 3D Shape Recognitioncited by 4
  5. 2022Composite Convolution: a Flexible Operator for Deep Learning on 3D Point Cloudscited by 4
  6. 2023Small but Mighty: Enhancing 3D Point Clouds Semantic Segmentation with U-Next Frameworkcited by 3