Object Detection Techniques Using Deep Learning: A Survey for Real-Time Applications

Abstract

Technologies in the field of autonomous vehicles (AV) has seen great evolution in recent years. Many automakers are actively attempting to embed advanced technologies into their products, and it’s testing at real world scenarios. This field of Autonomous Vehicles can arguably be said to be one of the most daunting topics of today in the field of Intelligent Transportation System (ITS), in particular the aspects like reliability, security, etc and as well as pushing forward for the world’s transition towards a highly sustainable future. The sensor technologies of today, however, have several drawbacks; wherein there are high levels of complexity and set-up costs likewise. Thus, this paper aims to accomplish the objective of object detection using only Computer Vision, and specifically for real-time purposes. Thus, this paper aims to explore and compare all the available deep learning based object detection algorithms and arrive at the best model for real-time applications.