Lightweight Vehicle Detection Based on Improved YOLOv5s
A vehicle detection algorithm is an indispensable component of intelligent traffic management and control systems, influencing the efficiency and functionality of the system. In this paper, we propose a lightweight improvement method for the YOLOv5 algorithm based on integrated perceptual attention,...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/4/1182 |
| _version_ | 1797297035420893184 |
|---|---|
| author | Yuhai Wang Shuobo Xu Peng Wang Kefeng Li Ze Song Quanfeng Zheng Yanshun Li Qiang He |
| author_facet | Yuhai Wang Shuobo Xu Peng Wang Kefeng Li Ze Song Quanfeng Zheng Yanshun Li Qiang He |
| author_sort | Yuhai Wang |
| collection | DOAJ |
| description | A vehicle detection algorithm is an indispensable component of intelligent traffic management and control systems, influencing the efficiency and functionality of the system. In this paper, we propose a lightweight improvement method for the YOLOv5 algorithm based on integrated perceptual attention, with few parameters and high detection accuracy. First, we propose a lightweight module IPA with a Transformer encoder based on integrated perceptual attention, which leads to a reduction in the number of parameters while capturing global dependencies for richer contextual information. Second, we propose a lightweight and efficient multiscale spatial channel reconstruction (MSCCR) module that does not increase parameter and computational complexity and facilitates representative feature learning. Finally, we incorporate the IPA module and the MSCCR module into the YOLOv5s backbone network to reduce model parameters and improve accuracy. The test results show that, compared with the original model, the model parameters decrease by about 9%, the average accuracy (mAP@50) increases by 3.1%, and the FLOPS does not increase. |
| first_indexed | 2024-03-07T22:15:24Z |
| format | Article |
| id | doaj.art-ddc26865e3f4434e81d71c621f5ed8ec |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-07T22:15:24Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-ddc26865e3f4434e81d71c621f5ed8ec2024-02-23T15:33:46ZengMDPI AGSensors1424-82202024-02-01244118210.3390/s24041182Lightweight Vehicle Detection Based on Improved YOLOv5sYuhai Wang0Shuobo Xu1Peng Wang2Kefeng Li3Ze Song4Quanfeng Zheng5Yanshun Li6Qiang He7School of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaSchool of Information and Electrical Engineering, Shandong Jiaotong University, Jinan 250357, ChinaA vehicle detection algorithm is an indispensable component of intelligent traffic management and control systems, influencing the efficiency and functionality of the system. In this paper, we propose a lightweight improvement method for the YOLOv5 algorithm based on integrated perceptual attention, with few parameters and high detection accuracy. First, we propose a lightweight module IPA with a Transformer encoder based on integrated perceptual attention, which leads to a reduction in the number of parameters while capturing global dependencies for richer contextual information. Second, we propose a lightweight and efficient multiscale spatial channel reconstruction (MSCCR) module that does not increase parameter and computational complexity and facilitates representative feature learning. Finally, we incorporate the IPA module and the MSCCR module into the YOLOv5s backbone network to reduce model parameters and improve accuracy. The test results show that, compared with the original model, the model parameters decrease by about 9%, the average accuracy (mAP@50) increases by 3.1%, and the FLOPS does not increase.https://www.mdpi.com/1424-8220/24/4/1182artificial intelligencedeep learningobject detectionvehicle detectionlightweight |
| spellingShingle | Yuhai Wang Shuobo Xu Peng Wang Kefeng Li Ze Song Quanfeng Zheng Yanshun Li Qiang He Lightweight Vehicle Detection Based on Improved YOLOv5s Sensors artificial intelligence deep learning object detection vehicle detection lightweight |
| title | Lightweight Vehicle Detection Based on Improved YOLOv5s |
| title_full | Lightweight Vehicle Detection Based on Improved YOLOv5s |
| title_fullStr | Lightweight Vehicle Detection Based on Improved YOLOv5s |
| title_full_unstemmed | Lightweight Vehicle Detection Based on Improved YOLOv5s |
| title_short | Lightweight Vehicle Detection Based on Improved YOLOv5s |
| title_sort | lightweight vehicle detection based on improved yolov5s |
| topic | artificial intelligence deep learning object detection vehicle detection lightweight |
| url | https://www.mdpi.com/1424-8220/24/4/1182 |
| work_keys_str_mv | AT yuhaiwang lightweightvehicledetectionbasedonimprovedyolov5s AT shuoboxu lightweightvehicledetectionbasedonimprovedyolov5s AT pengwang lightweightvehicledetectionbasedonimprovedyolov5s AT kefengli lightweightvehicledetectionbasedonimprovedyolov5s AT zesong lightweightvehicledetectionbasedonimprovedyolov5s AT quanfengzheng lightweightvehicledetectionbasedonimprovedyolov5s AT yanshunli lightweightvehicledetectionbasedonimprovedyolov5s AT qianghe lightweightvehicledetectionbasedonimprovedyolov5s |