An Axiomatic Analysis for Object Detection and Recognition Using Deep Learning

Abstract

AbstractIn the modern era, object detection and analysis is one of the crucial tasks. Numerous challenges have to be faced by the researchers to analyze the objects in images and real-time videos. Such analysis is required in multiple domains such as monitoring health, self-driving including detection of an anomaly, and many more. With the advanced usage of deep learning (DL) models and GPUs, the efficacy of object detection has been improved very well. Therefore, in this paper, we have provided a detailed study of both cases (images and videos) to detect specific objects using DL models. The existing approaches have been compared using on global wheat-head detection dataset to check the performance for object detection and recognition at different stages. From this analysis, we can say that DL is a backbone of computer vision to detect and recognize the various objects in a real-time scenario.KeywordsObject detectionGPUDeep learningYOLO