Eyes on the Road, Eyes on You: Next Generation Computer Vision for Improved Road Safety and Enhanced Driver Awareness

Our proposed project aims to revolutionize road safety and traffic management through the development of advanced computer vision algorithms for concurrent driver status monitoring and road traffic analysis. Leveraging a sophisticated dual-camera sensor setup, with one camera focused on external road conditions and the other on driver monitoring, our solution integrates a modified YOLO (You Only Look Once) object detection and classification algorithm. This modified YOLO algorithm is meticulously trained using rich datasets generated by our dual camera system. The external camera captures real-time data pertaining to traffic dynamics, vulnerable road users, and evolving road conditions. Simultaneously, the internal camera monitors the driver's behavior, enabling the algorithm to detect indicators of drowsiness, inattention, or incapacitation. By combining these insights, our system provides a holistic understanding of the driving environment, fostering a comprehensive approach to road safety. The key innovation lies in the adaptability of the YOLO algorithm, which is tailored to the unique requirements of our dual-camera system. This customization ensures optimized and accurate object detection and classification, enhancing the system's ability to identify not only common traffic elements but also specific driver-related parameters. Throughout this project, our focus will be on refining, training, and validating the algorithm using diverse datasets.