Pedestrian Intention Detection for Autonomous Driving: A Novel Object Detection Approach
A Padmavathi, Dheeraj Reddy Pullela, Nazeer Shaik · 2024
Detecting pedestrian intentions is crucial for safer interactions between autonomous driving and pedestrians. This study addresses this challenge by proposing an object detection model for autonomous driving using YOLOv8. Our model identifies pedestrians and categorizes their intentions as "waiting" or "crossing" based on bounding box analysis. Utilizing YOLOv8, the model recognizes pedestrians in real-time video streams and interprets their actions to determine their intentions. The model's performance is evaluated through metrics such as precision, recall, and F1 score, achieving high accuracy. This approach enables proactive decision-making based on pedestrian intentions, significantly enhancing autonomous driving and driver assistance system safety. The results highlight the model's potential to improve pedestrian safety in traffic, contribute to autonomous driving technology, and benefit connected vehicles.