[Poster Presentation]Improved YOLOv5 based on attention mechanism and FasterNet for foreign object detection on railway and airway tracks

Improved YOLOv5 based on attention mechanism and FasterNet for foreign object detection on railway and airway tracks
ID:67 Submission ID:261 View Protection:ATTENDEE Updated Time:2024-10-08 17:43:46 Hits:12 Poster Presentation

Start Time:2024-10-25 09:25 (Asia/Bangkok)

Duration:5min

Session:[PS] Poster Session » [PS] Poster

No files

Abstract
In recent years, there have been frequent incidents of foreign objects intruding into railway and Airport runways. These objects can include pedestrians, vehicles, animals, and debris. This paper introduces an improved YOLOv5 architecture incorporating FasterNet and attention mechanisms to enhance the detection of foreign objects on railways and Airport runways. This study proposes a new dataset, AARFOD (Aero and Rail Foreign Object Detection), which combines two public datasets for detecting foreign objects in aviation and railway systems. The dataset aims to improve the recognition capabilities of foreign object targets. Experimental results on this large dataset have demonstrated significant performance improvements of the proposed model over the baseline YOLOv5 model, reducing computational requirements. improved YOLO model shows a significant improvement in precision by 1.2%, recall rate by 1.0%, and mAP@.5 by 0.6%, while mAP@.5-.95 remained unchanged. The parameters were reduced by approximately 25.12%, and GFLOPs were reduced by about 10.63%. In the ablation experiment, it is found that the FasterNet module can significantly reduce the number of parameters of the model, and the reference of the attention mechanism can slow down the performance loss caused by lightweight.
 
Keywords
YOLOv5,FasterNet,NAM,foreign object detection
Speaker
Zongqing Qi
Computer Science Computer Science, Stevens Institute of Technology, Hoboken NJ, U.S

Submission Author
Zongqing Qi Computer Science, Stevens Institute of Technology, Hoboken NJ, U.S
Danqing Ma Computer Science, Stevens Institute of Technology, Hoboken NJ, U.S
Jingyu Xu Computer Information Technology, Northern Arizona University, Arizona, U.S
Ao Xiang Computer Science, University of Electronic Science and Technology of China, Sichuan, China
Hedi Qu Computer Science, Shenzhen SmartChip Microelectronics Technology Co., Ltd., China
Comment submit
Verification code Change another
All comments

CONTACT US

Conference Email: asiancomnet@usssociety.org

Whatsapp Group:  https://chat.whatsapp.com/HWRmX5hM1hFJKsbgMvpNTz

Meta(Facebook) Public Page: Usssociety.org

X(Twitter): @USSSOCIETY_ORG

 

 

Registration Submit Paper