English

Point Cloud Classification Using Content-based Transformer via Clustering in Feature Space

Computer Vision and Pattern Recognition 2023-07-21 v1

Abstract

Recently, there have been some attempts of Transformer in 3D point cloud classification. In order to reduce computations, most existing methods focus on local spatial attention, but ignore their content and fail to establish relationships between distant but relevant points. To overcome the limitation of local spatial attention, we propose a point content-based Transformer architecture, called PointConT for short. It exploits the locality of points in the feature space (content-based), which clusters the sampled points with similar features into the same class and computes the self-attention within each class, thus enabling an effective trade-off between capturing long-range dependencies and computational complexity. We further introduce an Inception feature aggregator for point cloud classification, which uses parallel structures to aggregate high-frequency and low-frequency information in each branch separately. Extensive experiments show that our PointConT model achieves a remarkable performance on point cloud shape classification. Especially, our method exhibits 90.3% Top-1 accuracy on the hardest setting of ScanObjectNN. Source code of this paper is available at https://github.com/yahuiliu99/PointConT.

Keywords

Cite

@article{arxiv.2303.04599,
  title  = {Point Cloud Classification Using Content-based Transformer via Clustering in Feature Space},
  author = {Yahui Liu and Bin Tian and Yisheng Lv and Lingxi Li and Feiyue Wang},
  journal= {arXiv preprint arXiv:2303.04599},
  year   = {2023}
}

Comments

This paper is accepted to IEEE/CAA Journal of Automatica Sinica (JAS)

R2 v1 2026-06-28T09:07:29.052Z