RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds

TL;DR

RandLA-Net is a lightweight neural network that efficiently performs semantic segmentation on large-scale 3D point clouds, achieving high accuracy and speed by using random sampling and a novel local feature aggregation module.

Contribution

We introduce RandLA-Net, a novel architecture that enables fast and accurate semantic segmentation of large-scale point clouds without complex sampling or pre/post-processing.

Findings

Processes 1 million points in a single pass

Up to 200X faster than existing methods

Surpasses state-of-the-art on Semantic3D and SemanticKITTI

Abstract

We study the problem of efficient semantic segmentation for large-scale 3D point clouds. By relying on expensive sampling techniques or computationally heavy pre/post-processing steps, most existing approaches are only able to be trained and operate over small-scale point clouds. In this paper, we introduce RandLA-Net, an efficient and lightweight neural architecture to directly infer per-point semantics for large-scale point clouds. The key to our approach is to use random point sampling instead of more complex point selection approaches. Although remarkably computation and memory efficient, random sampling can discard key features by chance. To overcome this, we introduce a novel local feature aggregation module to progressively increase the receptive field for each 3D point, thereby effectively preserving geometric details. Extensive experiments show that our RandLA-Net can process 1 million points in a single pass with up to 200X faster than existing approaches. Moreover, our RandLA-Net clearly surpasses state-of-the-art approaches for semantic segmentation on two large-scale benchmarks Semantic3D and SemanticKITTI.