"Object Detection Using Convolutional Neural Networks: A Review"

Abstract

The amount of data on the Internet has increased dramatically as a result of the advent of intelligent devices and social media. Object detection has become a popular international study topic as an important element of image processing. Convolutional Neural Network’s (CNN) remarkable capacity with feature learning and transfer learning has piqued attention in the computer vision field in recent years, resulting in a series of significant advancements in object identification. As a result, it's an important study on how to use CNN to improve object detection performance. The article began by explaining the core concept and architecture of CNN. Second, techniques for resolving current difficulties with traditional object detection are examined, with a focus on assessing detection algorithms based on region proposal and regression. Finally, it provided various methods for improving object detecting speed. The study then went on to discuss various publicly available object identification datasets as well as the notion of an assessment criterion. Finally, it went over existing object detection research results and ideas, highlighting significant advancements and outlining future prospects.