Recent Advances in Object Detection Based on YOLO‐V4 and Faster RCNN: A Review

Abstract

In 21st century object detection technique in computer vision and image processing is popular trends that deal with detecting occurrences of individuals, buildings, and vehicles, etc., from respective images and videos. Object Detection has received significant experimental interest in recent years due to its close relationship with video analysis and picture processing. For beginners, distinguishing between similar computer vision applications might take a lot of work. Currently, the approach to object recognition has grown into two categories: classic machine learning approaches that employ various computer vision methods and deep learning techniques that emerge from machine learning approaches for utilizing artificial neural networks. This article will introduce the subject of object identification as well as cutting-edge deep learning methods which are developed to handle it. We emphasize popular generic object detection models, as well as certain modifications and useful strategies for improving detection performance even further. For analysis, two significant types of deep learning approaches are chosen. The methods are FASTER R-CNN and YOLO-V4. The techniques presented in the article are briefly contrasted with the real-time images and videos for analyzing the objects.