TLD-READY: Traffic Light Detection ‐ Relevance Estimation and Deployment Analysis
Polley, Nikolai
1,2; Pavlitska, Svetlana 1; Boualili, Yacin 3; Rohrbeck, Patrick 3; Stiller, Paul 3; Bangaru, Ashok Kumar 3; Zöllner, J. Marius 1,2
1 Institut für Angewandte Informatik und Formale Beschreibungsverfahren (AIFB), Karlsruher Institut für Technologie (KIT)
2 Kompetenzzentrum für angewandte Sicherheitstechnologie (KASTEL), Karlsruher Institut für Technologie (KIT)
3 Karlsruher Institut für Technologie (KIT)
1,2; Pavlitska, Svetlana 1; Boualili, Yacin 3; Rohrbeck, Patrick 3; Stiller, Paul 3; Bangaru, Ashok Kumar 3; Zöllner, J. Marius 1,21 Institut für Angewandte Informatik und Formale Beschreibungsverfahren (AIFB), Karlsruher Institut für Technologie (KIT)
2 Kompetenzzentrum für angewandte Sicherheitstechnologie (KASTEL), Karlsruher Institut für Technologie (KIT)
3 Karlsruher Institut für Technologie (KIT)
Abstract:
Effective traffic light detection is a critical component of the perception stack in autonomous vehicles. This work introduces a novel deep-learning detection system while addressing the challenges of previous work.
Utilizing a comprehensive dataset amalgamation, including the Bosch Small Traffic Lights Dataset, LISA, the DriveU Traffic Light Dataset, and a proprietary dataset from Karlsruhe, we ensure a robust evaluation across varied scenarios.
Furthermore, we propose a relevance estimation system that innovatively uses directional arrow markings on the road, eliminating the need for prior map creation. On the DriveU dataset, this approach results in 96\% accuracy in relevance estimation. Finally, a real-world evaluation is performed to evaluate the deployment and generalizing abilities of these models.
For reproducibility and to facilitate further research, we provide the model weights and code: https://github.com/KASTEL-MobilityLab/traffic-light-detection
Utilizing a comprehensive dataset amalgamation, including the Bosch Small Traffic Lights Dataset, LISA, the DriveU Traffic Light Dataset, and a proprietary dataset from Karlsruhe, we ensure a robust evaluation across varied scenarios.
Furthermore, we propose a relevance estimation system that innovatively uses directional arrow markings on the road, eliminating the need for prior map creation. On the DriveU dataset, this approach results in 96\% accuracy in relevance estimation. Finally, a real-world evaluation is performed to evaluate the deployment and generalizing abilities of these models.
For reproducibility and to facilitate further research, we provide the model weights and code: https://github.com/KASTEL-MobilityLab/traffic-light-detection
| Zugehörige Institution(en) am KIT | Institut für Angewandte Informatik und Formale Beschreibungsverfahren (AIFB) Kompetenzzentrum für angewandte Sicherheitstechnologie (KASTEL) |
| Publikationstyp | Proceedingsbeitrag |
| Publikationsdatum | 24.09.2024 |
| Sprache | Englisch |
| Identifikator | ISBN: 979-83-315-0592-9 KITopen-ID: 1000182046 |
| HGF-Programm | 46.23.03 (POF IV, LK 01) Engineering Security for Mobility Systems |
| Erschienen in | 2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), Edmonton, 24th-27th September 2024 |
| Veranstaltung | 27th Internationla Conference on Intelligent Transportation Systems (ITSC 2024), Edmonton, Kanada, 24.09.2024 – 27.09.2024 |
| Verlag | Institute of Electrical and Electronics Engineers (IEEE) |
| Seiten | 3800–3806 |
| Schlagwörter | machine learning |
| Nachgewiesen in | OpenAlex Dimensions Scopus |
| Relationen in KITopen |
|
| Globale Ziele für nachhaltige Entwicklung |