Title: Safe-driving cloning by deep learning for autonomous cars

Authors: Wael Farag

Addresses: College of Engineering and Technology, American University of the Middle East, Kuwait; Electrical Engineering Department, Cairo University, Giza, Egypt

Abstract: In this paper, a convolutional neural network (CNN) to learn safe driving behaviour and smooth steering manoeuvring is proposed as an empowerment of autonomous driving technologies. The training data is collected from a front-facing camera and the steering commands issued by an experienced driver driving in traffic as well as urban roads. This data is then used to train the proposed CNN to facilitate what it is called 'behavioural cloning'. The proposed behaviour cloning CNN is named as 'BCNet', and its deep 17-layer architecture has been selected after extensive trials. The BCNet got trained using Adam's optimisation algorithm as a variant of the stochastic gradient descent (SGD) technique. The paper goes through the development and training process in details and shows the image processing pipeline harnessed in the development. The proposed approach is proved successful in cloning the driving behaviour embedded in the training dataset after extensive simulations.

Keywords: behavioural cloning; convolutional neural network; CNN; autonomous driving; machine learning.

DOI: 10.1504/IJAMECHS.2017.099318

International Journal of Advanced Mechatronic Systems, 2017 Vol.7 No.6, pp.390 - 397

Received: 06 Jul 2018
Accepted: 28 Nov 2018

Published online: 17 Apr 2019 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article