Deep Learning-Based Object Detection Approach for Autonomous Vehicles
Adnen Albouchi, Mohamed Ali Hajjaji, Abdellatif Mtibaa · 2022 IEEE 21st international Ccnference on Sciences and Techniques of Automatic Control and Computer Engineering (STA) · 2022
Recently, object detection is one of the most important methods in autonomous vehicles systems. To address the issue of false recognition and missed identification of small and blocked objects in autonomous vehicle situations, an enhanced deep learning object recognition approach is proposed. This paper proposes GPU-based implementation of intelligent multi-object detection and counting methods. In our approach, the pedestrian tracking by detection task is based on the 4Th version of YOLO and Deep SORT algorithms. For accurate multi-object detection and real-time support, an Enhanced one-stage YOLOv4 is presented. To monitor and count the discovered object, the deep simple online real-time tracker (SORT) method is used. The obtained results suggest that the proposed technique surpasses prior studies, indicating its efficacy as an intelligent vision-based approach for autonomous vehicles.