Adaptive Vision Framework for Low-Light Two-Wheeler Traffic Violation Detection Using Reinforcement-Aided YOLO-TVT

DOI

10.2991/978-94-6463-718-2_5How to use a DOI?

Keywords

Adaptive vision; low-light traffic monitoring; two-wheeler violations; YOLO-TVT; reinforcement learning; real-time detection; small object detection; modular framework; scalable traffic systems; privacy in traffic monitoring; smart city integration; automated traffic enforcement; low-light enhancement; traffic violation detection; real-time adaptation

Abstract

These two-wheeler traffic violations in low-light conditions are a major issue in traffic management and road safety. Current methods are mainly focused on one segment of the pipeline (e.g., object detection, low-light enhancement, or reinforcement learning) while failing to incorporate them into a single coherent system. The paper proposes Reinforcement-Aided YOLO-TVT, a novel Adaptive Vision Framework for Low-Light Two-Wheeler Traffic Violation Detection that addresses these limitations. This framework employs sophisticated low-light image enhancement methods along with a personalized YOLO architecture suited for detecting small objects like helmets and license plates, in poor lighting conditions. We incorporate reinforcement learning to allow real-time, adaptive decision-making to improve accuracy and reduce false positives. There are training on data from the perspectives of more than 2 years to Oct 2023. Deployment friendly system, hardware agnostic system which can easily work in low resource stringing environments. The proposed framework serves as a dynamic, privacy-preserving, intelligent solution for automated red-light traffic enforcement, equipped with rigorous privacy protections and capable of interfacing with smart city traffic management systems.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - V. Vennila
AU  - S. Savitha
AU  - A. Rajiv Kannan
AU  - B. Shanmathi
AU  - S. Syed Irfan
AU  - G. Vanmathi
PY  - 2025
DA  - 2025/05/23
TI  - Adaptive Vision Framework for Low-Light Two-Wheeler Traffic Violation Detection Using Reinforcement-Aided YOLO-TVT
BT  - Proceedings of the International Conference on Sustainability Innovation in Computing and Engineering (ICSICE 2024)
PB  - Atlantis Press
SP  - 37
EP  - 51
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-718-2_5
DO  - 10.2991/978-94-6463-718-2_5
ID  - Vennila2025
ER  -