Autonomous mini-buses are low-cost passenger vehicles that travel along designated routes in industrial parks. In order to achieve this goal, it is necessary to implement functionalities such as lane-keeping and obstacle avoidance. To address the challenge of deploying deep learning algorithms to detect environmental information on low-performance computing units, which leads to difficulties in model deployment and the inability to meet real-time requirements, a lightweight algorithm called YOLOP-E based on the YOLOP algorithm is proposed. (The letter ‘E’ stands for EfficientNetV2, and YOLOP-E represents the optimization of the entire algorithm by replacing the backbone of the original model with EfficientNetV2.) The algorithm has been optimized and improved in terms of the following three aspects: Firstly, the YOLOP backbone network is reconstructed using the lightweight backbone network EfficientNet-V2, and depth-wise separable convolutions are used instead of regular convolutions. Secondly, a hybrid attention mechanism called CABM is employed to enhance the model’s feature-representation capability. Finally, the Focal EIoU and Smoothed Cross-Entropy loss functions are utilized to improve detection accuracy. YOLOP-E is the final result after the aforementioned optimizations are completed. Experimental results demonstrate that on the BDD100K dataset, the optimized algorithm achieves a 3.5% increase in mAP50 and a 4.1% increase in mIoU. During real-world vehicle testing, the detection rate reaches 41.6 FPS, achieving the visual perception requirements of the autonomous shuttle bus while maintaining a lightweight design and improving detection accuracy.

More from our Archive

• DOI: 10.1177/10775463231188160 2023

  Modular quasi-zero-stiffness isolator based on compliant constant-force mechanisms for low-frequency vibration isolation

  Bingxiao Ding, Xuan Li, Shih-Chi Chen, Yangmin Li

• DOI: 10.3390/wevj14070177 2023

  Single-Wheel Failure Stability Control for Vehicle Equipped with Brake-by-Wire System

  Jian Zhou, Yage Di, Xuelong Miao

• DOI: 10.1177/10775463231185482 2023

  Adaptive backstepping dynamic sliding mode control for maglev turning motor spindle based on limited time preset performance

  Houcai Liu, Zheng Cao, Huimin Kang, Zhihai Ouyang, Lianghui Duan

• DOI: 10.1177/14680874231183768 2023

  Axis-symmetric diesel spray flame model coupled with momentum flux distribution measurement and improved soot phi-T map

  Takayuki Fuyuto, Yoshiyuki Mandokoro, Yuki Tagita, Tsutomu Umehara

• DOI: 10.1177/10775463231190263 2023

  Computational analysis of thermoelastic vibrations of functionally graded nonlocal nanobeam excited by thermal shock

  Ahmed E Abouelregal, Rakhi Tiwari

• DOI: 10.3390/wevj14070188 2023

  State of Health Assessment for Lithium-Ion Batteries Using Incremental Energy Analysis and Bidirectional Long Short-Term Memory

  Yanmei Li, Laijin Luo, Chaolong Zhang, Huihan Liu

• DOI: 10.1177/14680874231185421 2023

  Assessment of design and location of an active prechamber igniter to enable mixing-controlled combustion of ethanol in heavy-duty engines

  Jared Zeman, Ziming Yan, Michael Bunce, Adam Dempsey

• DOI: 10.1093/oxfclm/kgad005 2023

  The Paradox Test in Climate Litigation

  Adam Parr

• DOI: 10.3390/wevj14070179 2023

  Research on a Lightweight Panoramic Perception Algorithm for Electric Autonomous Mini-Buses

  Yulin Liu, Gang Li, Liguo Hao, Qiang Yang, Dong Zhang

• DOI: 10.4271/13-05-01-0006 2023

  Dynamics of Adopting Electric Vehicles in India: A Grounded Theory Approach

  Ankit Suri, B. Deepthi, Yogesh Sharma