remotesensing-11-02723.pdf (2.19 MB)
Download file

Learning control policies of driverless vehicles from UAV video streams in complex urban environments

Download (2.19 MB)
journal contribution
posted on 16.01.2020, 14:24 by Katie Inder, Varuna De-SilvaVaruna De-Silva, Xiyu ShiXiyu Shi
© 2019 by the authors. The way we drive, and the transport of today are going through radical changes. Intelligent mobility envisions to improve the e°ciency of traditional transportation through advanced digital technologies, such as robotics, artificial intelligence and Internet of Things. Central to the development of intelligent mobility technology is the emergence of connected autonomous vehicles (CAVs) where vehicles are capable of navigating environments autonomously. For this to be achieved, autonomous vehicles must be safe, trusted by passengers, and other drivers. However, it is practically impossible to train autonomous vehicles with all the possible tra°c conditions that they may encounter. The work in this paper presents an alternative solution of using infrastructure to aid CAVs to learn driving policies, specifically for complex junctions, which require local experience and knowledge to handle. The proposal is to learn safe driving policies through data-driven imitation learning of human-driven vehicles at a junction utilizing data captured from surveillance devices about vehicle movements at the junction. The proposed framework is demonstrated by processing video datasets captured from uncrewed aerial vehicles (UAVs) from three intersections around Europe which contain vehicle trajectories. An imitation learning algorithm based on long short-term memory (LSTM) neural network is proposed to learn and predict safe trajectories of vehicles. The proposed framework can be used for many purposes in intelligent mobility, such as augmenting the intelligent control algorithms in driverless vehicles, benchmarking driver behavior for insurance purposes, and for providing insights to city planning.

Funding

Engineering and Physical Sciences Research Council, grant number EP/T000783/1

History

School

  • Loughborough University London

Published in

Remote Sensing

Volume

11

Issue

23

Publisher

MDPI

Version

VoR (Version of Record)

Rights holder

© the Authors

Publisher statement

This is an Open Access Article. It is published by MDPI under the Creative Commons Attribution 4.0 Unported Licence (CC BY). Full details of this licence are available at: http://creativecommons.org/licenses/by/4.0/

Acceptance date

14/11/2019

Publication date

2019-11-20

Copyright date

2019

ISSN

2072-4292

eISSN

2072-4292

Language

en

Depositor

Dr Xiyu Shi Deposit date: 15 January 2020

Article number

2723