Multi-sensor fusion for enhanced object detection and tracking in autonomous driving.

Abstract

The development of autonomous vehicles is heavily reliant on the advancements in artificial intelligence (AI) and sensor fusion technologies. This project, a collaboration between the Cerist Research Center and the University of Bejaia, addresses the challenge of integrating data from multiple sensors to improve object detection and tracking accuracy. By leveraging state-of-the-art algorithms such as YOLOv8 for camera images and Complex YOLO for LiDAR point clouds, combined with the DeepSORT tracker and Kalman Filters, we propose a robust sensor fusion algorithm. Our solution has been evaluated using the KITTI dataset, showing significant improvements in tracking performance. The results demonstrate the potential of our approach to enhance the safety and reliability of autonomous vehicles.