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

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.

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

References in corpus (13)

• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• PVT v2: Improved Baselines with Pyramid Vision Transformer
• PCT: Point cloud transformer
• Transformer in Transformer
• PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies
• Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Framework
• Progressive LiDAR Adaptation for Road Detection
• How Do Vision Transformers Work?
• PRA-Net: Point Relation-Aware Network for 3D Point Cloud Analysis
• Shuffle Transformer: Rethinking Spatial Shuffle for Vision Transformer
• RangeSeg: Range-Aware Real Time Segmentation of 3D LiDAR Point Clouds
• BOAT: Bilateral Local Attention Vision Transformer
• Dynamic Group Transformer: A General Vision Transformer Backbone with Dynamic Group Attention

Cited by in corpus (2)

• Advancements in Point Cloud-Based 3D Defect Detection and Classification for Industrial Systems: A Comprehensive Survey
• Joint Learning for Scattered Point Cloud Understanding with Hierarchical Self-Distillation