1 article
-
Cyclist Safety Enhancement: A Multi-Modal Hazard Detection System
Abstract: This study presents a multi-modal hazard detection system to enhance cyclist safety in urban environments. Lever- aging a combination of computer vision, object tracking, and predictive modeling, the system offers a comprehensive approach to identifying and mitigating potential risks. Key contributions include improved depth estimation through object size priors, multi-class tracking utilizing KCF and Brisk, and a novel recurrent neural network architecture for predicting bicycle movement. The system’s collision detection …
Published in International Journal of Machine Systems and Manufacturing Technology · Vol. 1, Issue 2, 2023 · pp. 35–83 Read article