ADAS-PeVision: Advanced Driver Assistance System for Pedestrian Movement Prediction

With the advancement of autonomous vehicle innovation, precise pedestrian movement prediction has become indispensable for ensuring safe interactions between pedestrians and self-driving cars. In this research work, we propose a pedestrian tracking and prophesy framework intent at inflate pedestrian safety in densely populated environment. Our framework leverages state-of-the-art deep learning methodology, notably YOLO (You Only Look Once), for contemporaneous pedestrian detection and pursuit. By scrutinize the trajectories of detected pedestrians, we develop a prognosticative model to foretaste their future movements, thereby enabling self-driving cars to proactively adjust their etiquette to accustom pedestrian actions. To evaluate the efficacy of our approach, we conducted extensive demonstration using a diverse dataset of urban scenarios. Results demonstrate that our framework achieves high accuracy in prophesy pedestrian trajectories, with an average accuracy rate <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 1 \%}$</tex>. In addition to, our approach betrays robustness across various environmental circumstances such as sudden stops, direction changes, interaction with other objects in the environment and pedestrian behaviours. Overall, our research contributes to the advancement of pedestrian-aware autonomous navigation system, paving the way for safer and more methodical interactions between self-driving cars and pedestrians in urban environments. By enhancing pedestrian safety, our framework lays the footing for perceive the full potential of autonomous vehicles in transforming urban mobility and improving the quality of life for communities worldwide.