‘You Only Look Once’ Application for Autonomous Driving Vehicles & Cricket Spidercams using Convolutional Neural Network in Deep Learning

Abstract

Road safety is a prime concern in this era of high speed and automated driving vehicles. Lot of lives are lost or injured every day due to road accidents. Just understanding where the roads are is not adequate for an autonomous vehicle, obstacles like other vehicles and even less impact-resistant pedestrians and cyclists should be identified and avoided. Moreover, a technology proposed should also be capable to augment itself to provide other applications in the related fields. The proposed method here recognizes and report to the system about the objects such as cars, pedestrians, animals, etc. Once the object is identified, the next time vehicle approaches the similar object, it notifies the driver. And it also tells the system whether the object is moving towards or away from our vehicle. Augmenting this algorithm in applications like that of self-driven vehicle or automobiles/devices using Artificial Intelligence for the blind can be made for better safety. The system developed will be subjected to trials in the real life and correlated with an experimental setup.