A Probabilistic Formulation of LiDAR Mapping with Neural Radiance Fields

TL;DR

This paper introduces a probabilistic approach to training Neural Radiance Fields (NeRFs) for LiDAR data, enabling accurate scene reconstruction by modeling multiple returns and avoiding phantom surfaces.

Contribution

It proposes a novel probabilistic loss formulation for NeRFs that accounts for LiDAR's multiple returns, improving scene modeling accuracy.

Findings

Enhanced NeRF training with probabilistic loss reduces phantom surfaces.

Ability to model multiple LiDAR returns per ray improves scene fidelity.

Code availability facilitates adoption and further research.

Abstract

In this paper we reexamine the process through which a Neural Radiance Field (NeRF) can be trained to produce novel LiDAR views of a scene. Unlike image applications where camera pixels integrate light over time, LiDAR pulses arrive at specific times. As such, multiple LiDAR returns are possible for any given detector and the classification of these returns is inherently probabilistic. Applying a traditional NeRF training routine can result in the network learning phantom surfaces in free space between conflicting range measurements, similar to how floater aberrations may be produced by an image model. We show that by formulating loss as an integral of probability (rather than as an integral of optical density) the network can learn multiple peaks for a given ray, allowing the sampling of first, nth, or strongest returns from a single output channel. Code is available at https://github.com/mcdermatt/PLINK