Open Access"Object Detection Using Convolutional Neural Networks: A Review"
Author(s) -
Sushil Bhardwaj
Publication year - 2021
Publication title -
international journal of innovative research in computer science and technology
Language(s) - English
Resource type - Journals
ISSN - 2347-5552
DOI - 10.55524/ijircst.2021.9.6.64
Subject(s) - convolutional neural network , computer science , object detection , object (grammar) , artificial intelligence , field (mathematics) , identification (biology) , transfer of learning , machine learning , deep learning , cognitive neuroscience of visual object recognition , viola–jones object detection framework , feature (linguistics) , feature extraction , focus (optics) , pattern recognition (psychology) , face detection , linguistics , philosophy , botany , physics , mathematics , optics , pure mathematics , facial recognition system , biology
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.