Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control

While IL has been successfully applied in RL-based approaches for autonomous driving, significant challenges, such as limited data for RL and poor generalization in IL, still need further investigation. To overcome these limitations, we propose in this paper a novel approach that effectively combine...

Full description

Saved in:
Bibliographic Details
Main Authors: Badr Ben Elallid, Nabil Benamar, Miloud Bagaa, Sousso Kelouwani, Nabil Mrani
Format: Article
Language:English
Published: MDPI AG 2024-12-01
Series:World Electric Vehicle Journal
Subjects:
Online Access:https://www.mdpi.com/2032-6653/15/12/585
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846102346212311040
author Badr Ben Elallid
Nabil Benamar
Miloud Bagaa
Sousso Kelouwani
Nabil Mrani
author_facet Badr Ben Elallid
Nabil Benamar
Miloud Bagaa
Sousso Kelouwani
Nabil Mrani
author_sort Badr Ben Elallid
collection DOAJ
description While IL has been successfully applied in RL-based approaches for autonomous driving, significant challenges, such as limited data for RL and poor generalization in IL, still need further investigation. To overcome these limitations, we propose in this paper a novel approach that effectively combines IL with DRL by incorporating expert demonstration data to control AV in roundabout and right-turn intersection scenarios. Instead of employing CNNs, we integrate a ViT into the perception module of the SAC algorithm to extract key features from environmental images. The ViT algorithm excels in identifying relationships across different parts of an image, thereby enhancing environmental understanding, which leads to more accurate and precise decision making. Consequently, our approach not only boosts the performance of the DRL model but also accelerates its convergence, improving the overall efficiency and effectiveness of AVs in roundabouts and right-turn intersections with dense traffic by a achieving high success rate and low collision compared to RL baseline algorithms.
format Article
id doaj-art-5424823f31464c55b1c4bd47b9780ee8
institution Kabale University
issn 2032-6653
language English
publishDate 2024-12-01
publisher MDPI AG
record_format Article
series World Electric Vehicle Journal
spelling doaj-art-5424823f31464c55b1c4bd47b9780ee82024-12-27T14:59:38ZengMDPI AGWorld Electric Vehicle Journal2032-66532024-12-01151258510.3390/wevj15120585Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle ControlBadr Ben Elallid0Nabil Benamar1Miloud Bagaa2Sousso Kelouwani3Nabil Mrani4Computer Sciences at the School of Technology, Moulay Ismail University of Meknes, Meknes 50050, MoroccoComputer Sciences at the School of Technology, Moulay Ismail University of Meknes, Meknes 50050, MoroccoDepartment of Electrical and Computer Engineering, Université du Québec à Trois-Rivières, Québec City, QC G9A 5H7, CanadaDepartment of Electrical and Computer Engineering, Université du Québec à Trois-Rivières, Québec City, QC G9A 5H7, CanadaComputer Sciences at the School of Technology, Moulay Ismail University of Meknes, Meknes 50050, MoroccoWhile IL has been successfully applied in RL-based approaches for autonomous driving, significant challenges, such as limited data for RL and poor generalization in IL, still need further investigation. To overcome these limitations, we propose in this paper a novel approach that effectively combines IL with DRL by incorporating expert demonstration data to control AV in roundabout and right-turn intersection scenarios. Instead of employing CNNs, we integrate a ViT into the perception module of the SAC algorithm to extract key features from environmental images. The ViT algorithm excels in identifying relationships across different parts of an image, thereby enhancing environmental understanding, which leads to more accurate and precise decision making. Consequently, our approach not only boosts the performance of the DRL model but also accelerates its convergence, improving the overall efficiency and effectiveness of AVs in roundabouts and right-turn intersections with dense traffic by a achieving high success rate and low collision compared to RL baseline algorithms.https://www.mdpi.com/2032-6653/15/12/585autonomous drivingdeep reinforcement learningvision transformerimitation learning
spellingShingle Badr Ben Elallid
Nabil Benamar
Miloud Bagaa
Sousso Kelouwani
Nabil Mrani
Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control
World Electric Vehicle Journal
autonomous driving
deep reinforcement learning
vision transformer
imitation learning
title Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control
title_full Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control
title_fullStr Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control
title_full_unstemmed Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control
title_short Improving Reinforcement Learning with Expert Demonstrations and Vision Transformers for Autonomous Vehicle Control
title_sort improving reinforcement learning with expert demonstrations and vision transformers for autonomous vehicle control
topic autonomous driving
deep reinforcement learning
vision transformer
imitation learning
url https://www.mdpi.com/2032-6653/15/12/585
work_keys_str_mv AT badrbenelallid improvingreinforcementlearningwithexpertdemonstrationsandvisiontransformersforautonomousvehiclecontrol
AT nabilbenamar improvingreinforcementlearningwithexpertdemonstrationsandvisiontransformersforautonomousvehiclecontrol
AT miloudbagaa improvingreinforcementlearningwithexpertdemonstrationsandvisiontransformersforautonomousvehiclecontrol
AT soussokelouwani improvingreinforcementlearningwithexpertdemonstrationsandvisiontransformersforautonomousvehiclecontrol
AT nabilmrani improvingreinforcementlearningwithexpertdemonstrationsandvisiontransformersforautonomousvehiclecontrol