A hybrid object detection Approach for Autonomous vehicles using YOLOv8n model and CNN Transformer module

Autonomous vehicles are emerging day by day and it is being used as a part of public transport like taxi in some developed countries. The technology Edge artificial Intelligence is used for implementation of the models that running directly on local hardware devices instead of using remote cloud server. In this process sensor fusion, object detections, path finding are some of the important tasks. Object detection takes major role because, all objects in the road must be identified. Based on the observation, decision are made like whether the vehicle needs to be moved right or left or front or back etc. In this proposed research work the novel object detection approach is proposed to improve the accuracy of this process. YOLOv8n model and CNN Transformer module are combined and implemented to achieve the task. Based on the experimentation, it is observed that proposed model has produced the better results in terms of various Evaluation Metrics like mAP (mean Average Precision), Precision, Recall, F1-Score, Inference Time (FPS) .