Nighttime camera-based depth estimation is a highly challenging task, especially for autonomous driving applications, where accurate depth perception is essential for ensuring safe navigation. We aim to improve the reliability of perception systems at night time, where models trained on daytime data often fail in the absence of precise but costly LiDAR sensors. In this work, we introduce Light Enhanced Depth (LED), a novel cost-effective approach that significantly improves depth estimation in low-light environments by harnessing a pattern projected by high definition headlights available in modern vehicles. LED leads to significant performance boosts across multiple depth-estimation architectures (encoder-decoder, Adabins, DepthFormer) both on synthetic and real datasets. Furthermore, increased performances beyond illuminated areas reveal a holistic enhancement in scene understanding. Finally, we release the Nighttime Synthetic Drive Dataset, a new synthetic and photo-realistic nighttime dataset, which comprises 49,990 comprehensively annotated images.
@article{deMoreau2024led,title={LED: Light Enhanced Depth Estimation at Night},author={De Moreau, Simon and Almehio, Yasser and Bursuc, Andrei and El-Idrissi, Hafid and Stanciulescu, Bogdan and Moutarde, Fabien},journal={arXiv preprint arXiv:2409.08031},year={2024},}
2021
Development of agricultural robot platform with virtual laboratory capabilities
German
Monsalve, Oriane
Thiery, Simon
De Moreau, and
1 more author
In IECON 2021–47th Annual Conference of the IEEE Industrial Electronics Society, 2021
Agricultural robots are called to help in many tasks in emerging clean and sustainable agriculture. These complex electro-mechanical systems can actually integrate artificial intelligence (AI), the Internet of Things (IoT), sensors, actuators, and advanced control methods to accomplish functions in autonomous or in collaborative ways. Before the deployment of such techniques in the field, it is convenient to carry out laboratory validations. These last could be at the sub-system, e.g., sensors or servos operation, or the whole system level. This paper proposes the development of the hardware and software parts of a platform of agricultural robot. The proposed system, highly motivated by the restrictions imposed by COVID-19 context, enables laboratory tests virtualization while keeping real-time functionalities
@inproceedings{monsalve2021development,title={Development of agricultural robot platform with virtual laboratory capabilities},author={Monsalve, German and Thiery, Oriane and De Moreau, Simon and Cardenas, Alben},booktitle={IECON 2021--47th Annual Conference of the IEEE Industrial Electronics Society},pages={1--6},year={2021},organization={IEEE},}