LiDAS: Lighting-driven Dynamic Active Sensing for Nighttime Perception

Simon de Moreau1,2, Andrei Bursuc3, Hafid El-Idrissi3, Fabien Moutarde1,
1Mines Paris - PSL University, 2Valeo, 3valeo.ai
arXiv
LiDAS Overview

Abstract

Nighttime environments pose significant challenges for camera-based perception, as existing methods passively rely on the scene lighting. We introduce Lighting-driven Dynamic Active Sensing (LiDAS), a closed-loop active illumination system that combines off-the-shelf visual perception models with high‑definition headlights. Rather than uniformly brightening the scene, LiDAS dynamically predicts an optimal illumination field that maximizes downstream perception performance, i.e., decreasing light on empty areas to reallocate it on object regions. LiDAS enables zero-shot nighttime generalization of daytime-trained models through adaptive illumination control. Trained on synthetic data and deployed zero‑shot in real‑world closed‑loop driving scenarios, LiDAS enables +18.7% mAP50 and +5.0% mIoU over standard low‑beam at equal power. It maintains performances while reducing energy use by 40%. LiDAS complements domain‑generalization methods, further strengthening robustness without retraining. By turning readily available headlights into active vision actuators, LiDAS offers a cost‑effective solution to robust nighttime perception.

Video presentation

Qualitative results on synthetic dataset


Qualitative results

LiDAS lights objects of interest while leveraging ambient illumination, reducing power over the other car’s headlights and the pedestrian’s white coat. Only LiDAS detects the left-hand pedestrian and its segmentation map has well-defined objects and no sky artifacts.

Task-specific patterns


Novel datasets experiment
The downstream task used at training time directly shapes the learned lighting policy. LiDAS leverages both sparse, localized highlights over regions of interest from detection and broader, low-level illumination across the scene from segmentation when trained jointly.

Power-budget aware lighting


Qualitative results

LiDAS learns to prioritize the most informative regions under tight power budgets. As the budget grows, it progressively broadens coverage and begins to accentuate finer details that aid global scene understanding. Method[x] denotes power relative to LB.

BibTeX

@inproceedings{deMoreau2025lidas,
  title = {LiDAS: Lighting-driven Dynamic Active Sensing for Nighttime Perception},
  author = {De Moreau, Simon and Bursuc, Andrei and El-Idrissi, Hafid and Moutarde, Fabien},
  journal={arXiv preprint arXiv:xxxx},
  year={2025}
}