- Author
- Shanxing Ma (UGent) , Tim Willems (UGent) , Wenwen Ma (UGent) , Marwan Yusuf (UGent) , David Van Hamme (UGent) , Jan Aelterman (UGent) and Wilfried Philips (UGent)
- Organization
- Project
- Abstract
- As autonomous driving technology advances, the deployment of autonomous vehicles in urban environments is rapidly increasing. Lens flare—an often overlooked optical artifact in object detection research—can lead to increased false positives or missed detections, particularly in the challenging conditions inherent to autonomous driving. Current mitigation methods are often ill-suited for real-time implementation. This work proposes a solution to alleviate the adverse effects of lens flare by utilizing a lightweight lens flare perception network, eliminating the need for additional hardware or complex image pre-processing. Specifically, we propose a reference-free model utilizing a ResNet18 backbone integrated with a lightweight Multi-Layer Perceptron (MLP) to extract and leverage lens flare information. This model is developed via a teacher–student framework, which was distilled from an end-to-end reference-based model optimized using the Learned Perceptual Image Patch Similarity (LPIPS) metric. Our experiments demonstrate that incorporating lens flare information significantly enhances the performance of the baseline object detection network, outperforming previous mitigation methods by a substantial margin. The proposed method can be seamlessly integrated into existing object detectors and requires only an efficient training process, facilitating its deployment in practical autonomous driving tasks.
- Keywords
- autonomous driving, object detection, lens flare, likelihood ratio, OBJECT DETECTION, VEHICLES
Downloads
-
sensors-26-02359-v2.pdf
- full text (Published version)
- |
- open access
- |
- |
- 2.91 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-01KP875DW7Z85E3MN354GPKA7E
- MLA
- Ma, Shanxing, et al. “A Reference-Free Lens-Flare-Aware Detector for Autonomous Driving.” SENSORS, vol. 26, no. 8, 2026, doi:10.3390/s26082359.
- APA
- Ma, S., Willems, T., Ma, W., Yusuf, M., Van Hamme, D., Aelterman, J., & Philips, W. (2026). A reference-free lens-flare-aware detector for autonomous driving. SENSORS, 26(8). https://doi.org/10.3390/s26082359
- Chicago author-date
- Ma, Shanxing, Tim Willems, Wenwen Ma, Marwan Yusuf, David Van Hamme, Jan Aelterman, and Wilfried Philips. 2026. “A Reference-Free Lens-Flare-Aware Detector for Autonomous Driving.” SENSORS 26 (8). https://doi.org/10.3390/s26082359.
- Chicago author-date (all authors)
- Ma, Shanxing, Tim Willems, Wenwen Ma, Marwan Yusuf, David Van Hamme, Jan Aelterman, and Wilfried Philips. 2026. “A Reference-Free Lens-Flare-Aware Detector for Autonomous Driving.” SENSORS 26 (8). doi:10.3390/s26082359.
- Vancouver
- 1.Ma S, Willems T, Ma W, Yusuf M, Van Hamme D, Aelterman J, et al. A reference-free lens-flare-aware detector for autonomous driving. SENSORS. 2026;26(8).
- IEEE
- [1]S. Ma et al., “A reference-free lens-flare-aware detector for autonomous driving,” SENSORS, vol. 26, no. 8, 2026.
@article{01KP875DW7Z85E3MN354GPKA7E,
abstract = {{As autonomous driving technology advances, the deployment of autonomous vehicles in urban environments is rapidly increasing. Lens flare—an often overlooked optical artifact in object detection research—can lead to increased false positives or missed detections, particularly in the challenging conditions inherent to autonomous driving. Current mitigation methods are often ill-suited for real-time implementation. This work proposes a solution to alleviate the adverse effects of lens flare by utilizing a lightweight lens flare perception network, eliminating the need for additional hardware or complex image pre-processing. Specifically, we propose a reference-free model utilizing a ResNet18 backbone integrated with a lightweight Multi-Layer Perceptron (MLP) to extract and leverage lens flare information. This model is developed via a teacher–student framework, which was distilled from an end-to-end reference-based model optimized using the Learned Perceptual Image Patch Similarity (LPIPS) metric. Our experiments demonstrate that incorporating lens flare information significantly enhances the performance of the baseline object detection network, outperforming previous mitigation methods by a substantial margin. The proposed method can be seamlessly integrated into existing object detectors and requires only an efficient training process, facilitating its deployment in practical autonomous driving tasks.}},
articleno = {{2359}},
author = {{Ma, Shanxing and Willems, Tim and Ma, Wenwen and Yusuf, Marwan and Van Hamme, David and Aelterman, Jan and Philips, Wilfried}},
issn = {{1424-8220}},
journal = {{SENSORS}},
keywords = {{autonomous driving,object detection,lens flare,likelihood ratio,OBJECT DETECTION,VEHICLES}},
language = {{eng}},
number = {{8}},
pages = {{23}},
title = {{A reference-free lens-flare-aware detector for autonomous driving}},
url = {{http://doi.org/10.3390/s26082359}},
volume = {{26}},
year = {{2026}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: