FastPoint: Accelerating 3D Point Cloud Model Inference via Sample Point Distance Prediction

2025-07-31

AI Summary

Farthest Point Sampling (FPS) and neighborhood search incur substantial computational overhead in 3D point cloud model inference, hindering scalability to large-scale, irregular point clouds. Method: This work is the first to identify and model the temporal predictability of inter-point distance sequences during FPS. We propose a lightweight acceleration framework based on distance-trend prediction: instead of explicitly computing all pairwise distances, a deep learning model forecasts the evolution of distances, enabling dynamic pruning of redundant distance evaluations. Contribution/Results: Our method maintains sampling fidelity and downstream task accuracy while significantly accelerating the core preprocessing stage. On an NVIDIA RTX 3090, it achieves a 2.55× end-to-end inference speedup over baseline FPS. This establishes an efficient, practical paradigm for real-time, large-scale point cloud processing.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Distributed Search
• Machine Learning: Learning on the Edge & Model Compression

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Deep neural networks have revolutionized 3D point cloud processing, yet efficiently handling large and irregular point clouds remains challenging. To tackle this problem, we introduce FastPoint, a novel software-based acceleration technique that leverages the predictable distance trend between sampled points during farthest point sampling. By predicting the distance curve, we can efficiently identify subsequent sample points without exhaustively computing all pairwise distances. Our proposal substantially accelerates farthest point sampling and neighbor search operations while preserving sampling quality and model performance. By integrating FastPoint into state-of-the-art 3D point cloud models, we achieve 2.55x end-to-end speedup on NVIDIA RTX 3090 GPU without sacrificing accuracy.

Problem

• Accelerating 3D point cloud model inference
• Reducing computational cost of farthest point sampling
• Maintaining accuracy while speeding up neighbor search

Innovation

• Predicts distance curve for sample points
• Accelerates farthest point sampling efficiently
• Integrates without sacrificing model accuracy