Reproduce “Worsening Perception: Real-time Degradation of Autonomous Vehicle Perception Performance for Simulation of Adverse Weather Conditions” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ivan Fursa, Elias Fandi, Valentina Musat, Jacob Culley, Enric Gil, Izzeddin Teeti, Louise Bilous, Isaac Vander Sluis and 2 more
Researchers at Oxford Brookes University and collaborators devised a lightweight, real-time image augmentation method that accurately simulates autonomous vehicle perception degradation caused by adverse weather conditions. This approach quantitatively impacts object detection accuracy and facilitates repeatable, controlled testing for AV system robustness in both simulated and augmented reality environments.
Key insights
- Fading light conditions cause a significantly greater and more rapid degradation of autonomous vehicle perception performance compared to adherent water droplets on the camera lens.
- A focus on functionally degrading perception system outputs (e.g., mAP) is more critical for robust AV development than achieving photorealistic weather rendering in simulations.
- Lightweight, real-time augmentations can enable repeatable, controlled testing of the entire AV control pipeline in user-defined adverse conditions, applicable to both simulated and augmented real-world testing.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.