Artificial Intelligence based Object Detection System

Abstract

With applications in autonomous cars, smart surveillance, medical diagnostics, and industrial automation, object detection has emerged as a key component of computer vision and artificial intelligence. Deep learning has transformed object detection by increasing inference speed and accuracy. Three cutting-edge deep learning models—YOLOv8, Single Shot Multibook Detector (SSD), and Faster Region-based Convolutional Neural Network (Faster R-CNN)—are used in this extensive study to create an AI-based real-time object detection system. The ongoing advancement of object detection methods has led to the development of many systems during the last fifty years., it looks at the essential elements of creating a reliable real-time object detection pipeline, such as assessment metrics, training optimization, and dataset preparation. Additionally, it addresses new research trends like transformer-based architectures and lightweight deep learning networks, as well as the difficulties of real-time deployment on edge and low-power devices. The study comes to the conclusion that the future generation of intelligent visual perception systems can be powered by hybrid and adaptable models that combine the advantages of SSD, Faster R-CNN, and YOLOv8