Enhancing Pedestrian Safety in Autonomous Vehicles through Machine Learning

Autonomous Vehicles (AVs) have the potential to transform transportation by enhancing mobility, traffic efficiency, and road safety. Nevertheless, keeping pedestrians safe while AVs are around is a difficult task. The proposed work shows machine learning (ML) approaches can increase pedestrian safety in situations when autonomous vehicles are present. Using sensor data, computer vision and predictive modeling provide a comprehensive method for recognizing, anticipating, and reacting to pedestrian behaviors in real time. The introduction of ML models into AVs has the potential to lessen the risks presented by interactions between moving vehicles and people. It also contributes to the greater goal of creating more secure and harmonious urban transportation environments. The proposed work covers the development of machine learning models, real-time decision-making procedures, data collection and processing approaches, and preliminary findings stressing the potential of ML-powered AVs to reduce pedestrian-vehicle accidents. Furthermore, through simulation and real-world testing, the research provides persuasive evidence of the system's usefulness. The simulation results show that ML-enhanced systems outperform the previous system in terms of pedestrian identification, prediction, and collision avoidance rates. Real-world testing validates the improved system's increased performance, demonstrating that it has a substantial potential to improve pedestrian safety in autonomous car scenarios.