A Review on Deep Learning Based Object Detection Models
Authors: Nazir, A., Wani, M.A.
Journal: Proceedings of the 2026 13th International Conference on Computing for Sustainable Global Development Indiacom 2026
Publication Date: 01/01/2026
DOI: 10.23919/INDIACom70271.2026.11526524
Abstract:Object detection, using deep learning approaches, has added a new dimension to the field of computer vision. The goal of this study is to analyse the architectural layout, performance and applications of one-stage and two-stage object detection frameworks. The most ubiquitous object detection using deep learning approaches rely on convolutional neural networks (CNNs), which are trained on huge datasets of annotated images. Object detection aims to identify various occurrences of items either small or large belonging to various predetermined classes within images of the real world. This is one of the most challenging problems to find small instances of objects in computer vision. In this review, we spotlight on the five two-stage region proposal-based object identification methods: R-CNN, Faster R-CNN, Fast R-CNN, Mask R-CNN, and seven single-stage YOLO variants for object detection. The architectures of object detection models are analyzed based on several key aspects, including their backbone network, region proposal mechanism, and prediction heads. Also, the utilization of object detection methods using deep learning is assessed in relation to their application in fields like self-driving cars, security monitoring systems, and medical sciences.
Source: Scopus