SLiDE: Self-supervised LiDAR De-snowing through Reconstruction Difficulty

TL;DR

This paper introduces a self-supervised learning framework for removing snow noise from LiDAR point clouds, leveraging the structural properties of noise points without requiring manual annotations.

Contribution

The proposed method uses two neural networks to detect snow points based on reconstruction difficulty, achieving state-of-the-art label-free performance and enhancing supervised de-snowing training.

Findings

Achieves state-of-the-art results among label-free methods.

Comparable performance to fully-supervised approaches.

Improves label efficiency in supervised training.

Abstract

LiDAR is widely used to capture accurate 3D outdoor scene structures. However, LiDAR produces many undesirable noise points in snowy weather, which hamper analyzing meaningful 3D scene structures. Semantic segmentation with snow labels would be a straightforward solution for removing them, but it requires laborious point-wise annotation. To address this problem, we propose a novel self-supervised learning framework for snow points removal in LiDAR point clouds. Our method exploits the structural characteristic of the noise points: low spatial correlation with their neighbors. Our method consists of two deep neural networks: Point Reconstruction Network (PR-Net) reconstructs each point from its neighbors; Reconstruction Difficulty Network (RD-Net) predicts point-wise difficulty of the reconstruction by PR-Net, which we call reconstruction difficulty. With simple post-processing, our method effectively detects snow points without any label. Our method achieves the state-of-the-art performance among label-free approaches and is comparable to the fully-supervised method. Moreover, we demonstrate that our method can be exploited as a pretext task to improve label-efficiency of supervised training of de-snowing.