A Reinforcement Learning Framework for Video Frame-Based Autonomous Car-Following

Abstract

Car-following theory has received considerable attention as a core component of Intelligent Transportation Systems. However, its application to the emerging autonomous vehicles (AVs) remains an unexplored research area. AVs are designed to provide convenient and safe driving by avoiding accidents caused by human errors. They require advanced levels of recognition of other drivers' driving-style. With car-following models, AVs can use their built-in technology to understand the environment surrounding them and make real-time decisions to follow other vehicles. In this paper, we design an end-to-end carfollowing framework for AVs using automated object detection and navigation decision modules. The objective is to allow an AV to follow another vehicle based on Red Green Blue Depth (RGB-D) frames. We propose to employ a joint solution involving the You Look Once version 3 (YOLOv3) object detector to identify the leader vehicle and other obstacles and a reinforcement learning (RL) algorithm to navigate the self-driving vehicle. Two RL algorithms, namely Q-learning and Deep Q-learning have been investigated. Simulation results show the convergence of the developed models and investigate their efficiency in following the leader. It is shown that, with video frames only, promising results are achieved and that AVs can adopt a reasonable car-following behavior.

A Machine Learning Smartphone-based Sensing for Driver Behavior Classification

“…After modeling multiple driving behaviors and generating different scenarios of trips, we collect more than 500 minutes using the Carla simulator as shown in Fig. 2 [16]- [18]. The driving sessions are recorded on four different towns with two types of road (motorway and secondary) with four different weather conditions (sunny, soft rain, cloudy & foggy, and stormy) and four styles of driving (normal, intermediate, aggressive, and dangerous) that occur along the trips such that a reasonable balanced dataset between the different classes is generated.…”

Deep Learning Control for Autonomous Robot

“…The advantages of self-driving vehicles include the ability to predict and manage traffic problems and provide built-up accessibility for all users. Academic and industry research groups have developed various low-cost systems for studying research problems in autonomous driving and control [7][8][9][10][11][12].…”

End-to-End Neural Network for Autonomous Steering using LiDAR Point Cloud Data

“…The typical driving automation technology is generally a rule-based method, which obtains environmental information through on-board sensors and develops automated driving rules based on those of human driving [1], [2]. The vehicle completes the autonomous driving control according to the expert rule system.…”