iju123/AL_Assisted_Traffic_light_Control_system

🚦 AI-Assisted Traffic Light Control System

This is a real-time AI traffic light controller built with YOLOv8 + PyTorch + Gradio. It detects cars, trucks, and ambulances from camera feeds and gives priority to ambulances in traffic.

✨ Features

• —Detects cars, trucks, ambulances from live video/images
• —If an ambulance is detected → traffic light turns GREEN ✅
• —Trucks can be double-checked with an optional ambulance classifier
• —Built with Gradio for easy deployment and usage
• —Runs on Hugging Face Spaces (no setup needed)

🛠️ How It Works

1. 1.YOLOv8 model detects vehicles in the frame.
2. 2.If YOLO labels a vehicle as a truck, we optionally pass it to a small CNN classifier:
3. 3.ambulance_classifier.pth → determines if it's an ambulance.
4. 4.If missing → fallback to YOLO only.
5. 5.Traffic rules (`traffic_logic.py`):
6. 6.Ambulance → 🚦 GREEN LIGHT (priority passage)
7. 7.Otherwise → 🚦 Normal rules for cars & trucks.

🚀 Run Locally

Clone the repo and install dependencies:

git clone https://huggingface.co/spaces/your-username/traffic-light-ai
cd traffic-light-ai
pip install -r requirements.txt