Year: 2025
Type: article
Abstract: - Traffic congestion is a critical challenge in urban areas, leading to increased travel time, fuel consumption, environmental pollution, and economic losses. Traditional traffic control systems rely on fixed signal timings that fail to adapt to real-time traffic conditions, resulting in inefficiencies. To address this issue, we introduce SmartFlow, an advanced adaptive traffic management system powered by YOLOv8, a state-of-the-art real-time object detection model. SmartFlow leverages a hybrid deep learning framework that integrates active learning and network adaptation to optimize traffic signal control dynamically.The SmartFlow system utilizes existing CCTV infrastructure to monitor vehicular density at intersections, eliminating the need for costly physical sensors. By analyzing real-time traffic patterns, it adjusts signal durations dynamically, ensuring smoother traffic flow and reduced congestion. A key feature of SmartFlow is its emergency vehicle prioritization mechanism, which detects ambulances and fire trucks, allowing them to navigate through intersections with minimal delay, thereby improving emergency response times.To enhance accuracy and efficiency, SmartFlow incorporates attention-based optimization and reinforcement learning strategies. The attention module enables the system to focus on critical traffic parameters, minimizing inefficiencies in signal adjustments. Key Words: Traffic congestion, Adaptive traffic control, YOLOv8, Real-time object detection, Active learning, Network adaptation, Smart traffic management, Emergency vehicle prioritization, CCTV-based traffic monitoring, Intelligent transportation systems
DOI: 10.55041/ijsrem48211
Author: Mohamed Sameer I
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4410476970