Traffic congestion in urban areas significantly delays emergency vehicles like an ambulance, increasing the risk of fatal outcomes. This paper presents an IoT and deep learning-based traffic control system that prioritizes ambulances at intersections. The system utilizes the YOLOv8 object detection model to detect ambulances accurately in real-time from camera feeds. Upon detection of an ambulance, the system processes the data via a web application created in Python Flask and interfaced with an Arduino microcontroller to switch the traffic lights to allow unhindered passage for the ambulance. The proposed solution can provide higher efficiency for emergency response by reducing signaling waiting time and integrating deep learning and IoT with real-time traffic control. Good detection accuracy and fast signal responsiveness were demonstrated in preliminary testing. This research can help advance smart city infrastructure by providing an automated and scalable traffic control system for emergency vehicles.