東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Intelligent Intersection Navigation ...

Sassi, Ayoub,

FindBook

Google Book

Amazon

博客來

Intelligent Intersection Navigation Enhancing Autonomous Driving With Reinforcement Learning and Bird's Eye View Fusion /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Intelligent Intersection Navigation Enhancing Autonomous Driving With Reinforcement Learning and Bird's Eye View Fusion // Ayoub Sassi.
作者:	Sassi, Ayoub,
面頁冊數:	1 electronic resource (62 pages)
附註:	Source: Masters Abstracts International, Volume: 87-11.
Contained By:	Masters Abstracts International87-11.
標題:	Automotive engineering. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=32668563
ISBN:	9798241691668

Intelligent Intersection Navigation Enhancing Autonomous Driving With Reinforcement Learning and Bird's Eye View Fusion /
Sassi, Ayoub,

Intelligent Intersection Navigation Enhancing Autonomous Driving With Reinforcement Learning and Bird's Eye View Fusion /Ayoub Sassi. - 1 electronic resource (62 pages)

Source: Masters Abstracts International, Volume: 87-11.

In recent years, the automotive sector has witnessed significant advancements driven by the integration of Artificial Intelligence (AI). Autonomous driving systems are reshaping modern transportation by improving safety, efficiency, and accessibility. However, handling complex traffic situations remains a major challenge due to their dynamic and uncertain nature. Conventional rule-based approaches often lack the flexibility required to manage such environments, highlighting the need for more adaptive and intelligent solutions. In this work, we address the problem of autonomous navigation at intersections, one of the most challenging scenarios in urban driving. We introduce a novel approach that combines Bird's Eye View (BEV) representation with Reinforcement Learning (RL) to enhance decision-making capabilities. Using the CARLA simulator as a controlled environment for training and evaluation, we first fine-tune the UNetXST model to generate a fused BEV representation from multiple camera views. This unified perspective provides a comprehensive understanding of the surrounding environment and serves as input to the learning agent. Then, we employ the Proximal Policy Optimization (PPO) algorithm to train an RL agent capable of learning effective driving strategies through continuous interaction with the simulated environment. The objective is to achieve safe and efficient intersection traversal while minimizing collisions and ensuring smooth traffic behavior. This work contributes toward the development of more robust autonomous driving systems, with potential implications for enhancing road safety and optimizing urban mobility.

English

ISBN: 9798241691668Subjects--Topical Terms:

2181195
Automotive engineering.
Subjects--Index Terms:

Autonomous driving systems

Intelligent Intersection Navigation Enhancing Autonomous Driving With Reinforcement Learning and Bird's Eye View Fusion /
LDR:03133nmm a22004333i 4500 001 2427668
005 20260911085220.5
006 m o d
007 cr|nu|||||||||
008 260929s2026 miu||||||m |||||||eng d
020 $a 9798241691668
035 $a (MiAaPQD)AAI32668563
035 $a AAI32668563
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Sassi, Ayoub, $e author. $3 3814667
245 1 0 $a Intelligent Intersection Navigation Enhancing Autonomous Driving With Reinforcement Learning and Bird's Eye View Fusion / $c Ayoub Sassi.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2026
300 $a 1 electronic resource (62 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 87-11.
500 $a Advisors: Zedini, Emna Committee members: Mkaouer, Mohamed Wiem; Soui, Makram.
502 $b M.S. $c University of Michigan-Flint $d 2026.
520 $a In recent years, the automotive sector has witnessed significant advancements driven by the integration of Artificial Intelligence (AI). Autonomous driving systems are reshaping modern transportation by improving safety, efficiency, and accessibility. However, handling complex traffic situations remains a major challenge due to their dynamic and uncertain nature. Conventional rule-based approaches often lack the flexibility required to manage such environments, highlighting the need for more adaptive and intelligent solutions. In this work, we address the problem of autonomous navigation at intersections, one of the most challenging scenarios in urban driving. We introduce a novel approach that combines Bird's Eye View (BEV) representation with Reinforcement Learning (RL) to enhance decision-making capabilities. Using the CARLA simulator as a controlled environment for training and evaluation, we first fine-tune the UNetXST model to generate a fused BEV representation from multiple camera views. This unified perspective provides a comprehensive understanding of the surrounding environment and serves as input to the learning agent. Then, we employ the Proximal Policy Optimization (PPO) algorithm to train an RL agent capable of learning effective driving strategies through continuous interaction with the simulated environment. The objective is to achieve safe and efficient intersection traversal while minimizing collisions and ensuring smooth traffic behavior. This work contributes toward the development of more robust autonomous driving systems, with potential implications for enhancing road safety and optimizing urban mobility.
546 $a English
590 $a School code: 1873
650 4 $a Automotive engineering. $3 2181195
650 4 $a Computer science. $3 523869
653 $a Autonomous driving systems
653 $a Bird's Eye View
653 $a Intersection traversal
653 $a Proximal Policy Optimization
653 $a Reinforcement Learning
690 $a 0800
690 $a 0540
690 $a 0984
710 2 $a University of Michigan-Flint. $b Computer Science & Information Systems. $e degree granting institution. $3 3814668
720 1 $a Zedini, Emna $e degree supervisor.
773 0 $t Masters Abstracts International $g 87-11.
790 $a 1873
791 $a M.S.
792 $a 2026
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=32668563