Lid-Lab-NeRF: Generating Temporally Consistent, Labelled LiDAR Point Clouds using Neural Radiance Fields


Shrestha Srivastava (Indian Institute of Science Education and Research Bhopal), Vaibhav Kumar (Indian Institute of Science Education and Research Bhopal)
The 35th British Machine Vision Conference
PDFPosterVideo (Right click to download)Supplementary

Abstract

LiDAR point clouds are essential for autonomous driving and urban planning, but collecting and labelling high-quality data remains expensive and inflexible due to fixed vehicle trajectories. While synthetic data generation offers a potential solution, existing approaches struggle to bridge the reality gap and often produce temporally inconsistent or unlabelled results. We present Lid-Lab-NeRF, a novel framework that generates labelled, dynamic LiDAR point clouds by combining neural radiance fields with motion field prediction for better time-consistent outputs and semantic labelling through a semantic radiance field. Our model simultaneously produces intensity, depth, labels, and raydrop probability, while a feature alignment loss and raydrop optimisation pipeline ensure realistic scan patterns. Experiments on the KITTI-360 dataset demonstrate that Lid-Lab-NeRF effectively synthesizes labelled point clouds that closely match real LiDAR characteristics compared to existing works. Codes are available at https://github.com/geoai4cities/Lid-Lab-NeRF.git

Citation

@inproceedings{Srivastava_2025_BMVC,
author    = {Shrestha Srivastava and Vaibhav Kumar},
title     = {Lid-Lab-NeRF: Generating Temporally Consistent, Labelled LiDAR Point Clouds using Neural Radiance Fields},
booktitle = {36th British Machine Vision Conference 2025, {BMVC} 2025, Sheffield, UK, November 24-27, 2025},
publisher = {BMVA},
year      = {2025},
url       = {https://bmva-archive.org.uk/bmvc/2025/assets/papers/Paper_999/paper.pdf}
}


Copyright © 2025 The British Machine Vision Association and Society for Pattern Recognition
The British Machine Vision Conference is organised by The British Machine Vision Association and Society for Pattern Recognition. The Association is a Company limited by guarantee, No.2543446, and a non-profit-making body, registered in England and Wales as Charity No.1002307 (Registered Office: Dept. of Computer Science, Durham University, South Road, Durham, DH1 3LE, UK).

Imprint | Data Protection