PCT: Point cloud transformer

2021/04/09 by Meng-Hao Guo, Jun-Xiong Cai, Zheng-Ning Liu +3 · 27 citations
Engineering · Earth and Planetary Sciences · Computer Science · #3D Shape Modeling and Analysis #3D Surveying and Cultural Heritage #Optical measurement and interference techniques

paper · pdf · doi:10.1007/s41095-021-0229-5

openalex created_date 2020/12/21 · openalex publication_date 2021/04/09 · openalex updated_date 2026/07/31

Abstract

The irregular domain and lack of ordering make it challenging to design deep neural networks for point cloud processing. This paper presents a novel framework named Point Cloud Transformer (PCT) for point cloud learning. PCT is based on Transformer, which achieves huge success in natural language processing and displays great potential in image processing. It is inherently permutation invariant for processing a sequence of points, making it well-suited for point cloud learning. To better capture local context within the point cloud, we enhance input embedding with the support of farthest point sampling and nearest neighbor search. Extensive experiments demonstrate that the PCT achieves the state-of-the-art performance on shape classification, part segmentation, semantic segmentation, and normal estimation tasks.

Cited by

• DAP-MAE: Domain-Adaptive Point Cloud Masked Autoencoder for Effective Cross-Domain Learning
• GauSSmart: Enhanced 3D Reconstruction through 2D Foundation Models and Geometric Filtering
• MambaH-Fit: Rethinking Hyper-surface Fitting-based Point Cloud Normal Estimation via State Space Modelling
• Vision-Language-Action Models for Robotics: A Review Towards Real-World Applications
• TACO-Net: Topological Signatures Triumph in 3D Object Classification
• Neural Visibility of Point Sets
• Desensitizing for Improving Corruption Robustness in Point Cloud Classification through Adversarial Training
• Wavelet-Induced Rotary Encodings: RoPE Meets Graphs
• Point-RTD: Replaced Token Denoising for Pretraining Transformer Models on Point Clouds
• Unlocking Hidden Potential in Point Cloud Networks with Attention-Guided Grouping-Feature Coordination
• EMPEROR: Efficient Moment-Preserving Representation of Distributions
• Hierarchical Self-Attention: Generalizing Neural Attention Mechanics to Multi-Scale Problems
• WHU-STree: A Multi-modal Benchmark Dataset for Street Tree Inventory
• VortexTransformer: End‐to‐End Objective Vortex Detection in 2D Unsteady Flow Using Transformers
• Enhancing 3D Point Cloud Classification with ModelNet-R and Point-SkipNet
• Selective Structural Ablation for Efficient 3D Point Cloud Signal Processing
• Towards More Diverse and Challenging Pre-training for Point Cloud Learning: Self-Supervised Cross Reconstruction with Decoupled Views
• Adaptive Point-Prompt Tuning: Fine-Tuning Heterogeneous Foundation Models for 3D Point Cloud Analysis
• Towards a 3D Transfer-based Black-box Attack via Critical Feature Guidance
• TCFNet: Bidirectional face-bone transformation via a Transformer-based coarse-to-fine point movement network
• Masked Clustering Prediction for Unsupervised Point Cloud Pre-training
• Open-world Point Cloud Semantic Segmentation: A Human-in-the-loop Framework
• HydraMamba: Multi-Head State Space Model for Global Point Cloud Learning
• Generating Adversarial Point Clouds Using Diffusion Model
• Transferable and Undefendable Point Cloud Attacks via Medial Axis Transform
• 3D Test-time Adaptation via Graph Spectral Driven Point Shift
• HybridTM: Combining Transformer and Mamba for 3D Semantic Segmentation

Related

• Point Cloud Learning with Transformer
• Dual Transformer for Point Cloud Analysis
• CpT: Convolutional Point Transformer for 3D Point Cloud Processing
• Adaptive Point Transformer
• Cloud Transformers: A Universal Approach To Point Cloud Processing Tasks