Comparative Analysis of Yolo and SSD

Abstract

Object detection is a computer vision technique. It plays a significant role in healthcare, autonomous vehicles, facial recognition and traffic management system, license plate recognition and so on. Our paper will focus on the existing deep learning techniques based on convolutional neural networks and how these techniques are different from each other. These techniques include single stage models such as -YOLOV7, YOLOV8, and SSD. Our paper will compare these techniques to draw meaningful conclusions. With the help of these single stage models, we are implementing vehicle detection systems and then comparing these model's performance. In comparative analysis, we are using Coco dataset, we evaluated the performance of these techniques based on parameters such as: mean average precision, inference time and recall time.