# Vision-Link AI: Real-time Accident Detection and Safety Analysis using YOLO11

‎Overview: This report presents Vision-Link AI, a deep learning framework developed for the ACCIDENT @ CVPR 2026 competition. By leveraging the YOLO11 architecture, we achieved a remarkable processing speed of 7.0ms per frame. Our model demonstrates high reliability in detecting vehicle collisions across diverse environments, including night-time and high-speed highway scenarios.
‎
‎Technical Paper:
‎You can access our full 5-page research report via the Google Drive link in the attachments below.
[https://drive.google.com/file/d/1I25L5NmeiedFCQ8T7HHrRj84SiPsHtVA/view?usp=drivesdk](https://drive.google.com/file/d/1I25L5NmeiedFCQ8T7HHrRj84SiPsHtVA/view?usp=drivesdk)