Accident Detection Using AI (YOLOv8)

A real-time accident detection system built using YOLOv8, capable of identifying road collisions, falls, and hazardous events from live video feeds with high accuracy. The project includes custom dataset preparation, training, and model optimization for low-latency inference on edge devices. The system processes CCTV or dash-cam streams, detects accidents instantly, and triggers automated alerts through a Python backend—showcasing strong skills in computer vision, deep learning, and real-time video analytics.

Core Features:
– Real-time accident detection using YOLOv8 with high frame-rate inference.
– Custom-trained model for recognizing collisions, falls, and hazardous events.
– Robust performance across different lighting, weather, and traffic conditions.

Files Included: