Improved coordinate attention network for classification of dangerous driving behavior
With the rise of traffic accidents caused by unsafe driving behaviors, the accurate classification of these behaviors has become a pressing issue in intelligent transportation systems. Traditional methods such as AlexNet and VGG, while effective for general image recognition tasks, fail to capture t...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Franklin Open |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S277318632500009X |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832542425910870016 |
|---|---|
| author | Wen Ni Lufeng Bai |
| author_facet | Wen Ni Lufeng Bai |
| author_sort | Wen Ni |
| collection | DOAJ |
| description | With the rise of traffic accidents caused by unsafe driving behaviors, the accurate classification of these behaviors has become a pressing issue in intelligent transportation systems. Traditional methods such as AlexNet and VGG, while effective for general image recognition tasks, fail to capture the complex and subtle features necessary for recognizing dangerous driving behaviors. To address this, we propose an improved residual network model, SC-ResNet, which integrates a coordinate attention mechanism and SIFT (Scale-Invariant Feature Transform) feature fusion to enhance classification accuracy under varying conditions including rotation, scale, and illumination changes. Furthermore, we introduce a multi-scale feature pyramid network and a novel joint loss function to better handle the multi-class classification imbalance problem. Experimental results show that our model outperforms traditional networks by 0.6% to 4.7% in classification accuracy. Future research will focus on improving model generalization and computational efficiency for real-time applications. |
| format | Article |
| id | doaj-art-1ac1735d834e484bafd6c8de4e2d604d |
| institution | Kabale University |
| issn | 2773-1863 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Franklin Open |
| spelling | doaj-art-1ac1735d834e484bafd6c8de4e2d604d2025-02-04T04:10:43ZengElsevierFranklin Open2773-18632025-03-0110100219Improved coordinate attention network for classification of dangerous driving behaviorWen Ni0Lufeng Bai1Jiangsu Second Normal University Nanjing, Jiangsu 211200, China; School of Economics and Management, University of Science and Technology, Nanjing, 210094, ChinaJiangsu Second Normal University Nanjing, Jiangsu 211200, China; Corresponding author.With the rise of traffic accidents caused by unsafe driving behaviors, the accurate classification of these behaviors has become a pressing issue in intelligent transportation systems. Traditional methods such as AlexNet and VGG, while effective for general image recognition tasks, fail to capture the complex and subtle features necessary for recognizing dangerous driving behaviors. To address this, we propose an improved residual network model, SC-ResNet, which integrates a coordinate attention mechanism and SIFT (Scale-Invariant Feature Transform) feature fusion to enhance classification accuracy under varying conditions including rotation, scale, and illumination changes. Furthermore, we introduce a multi-scale feature pyramid network and a novel joint loss function to better handle the multi-class classification imbalance problem. Experimental results show that our model outperforms traditional networks by 0.6% to 4.7% in classification accuracy. Future research will focus on improving model generalization and computational efficiency for real-time applications.http://www.sciencedirect.com/science/article/pii/S277318632500009XCoordinate attentionDangerous drivingExponential cross entropy |
| spellingShingle | Wen Ni Lufeng Bai Improved coordinate attention network for classification of dangerous driving behavior Franklin Open Coordinate attention Dangerous driving Exponential cross entropy |
| title | Improved coordinate attention network for classification of dangerous driving behavior |
| title_full | Improved coordinate attention network for classification of dangerous driving behavior |
| title_fullStr | Improved coordinate attention network for classification of dangerous driving behavior |
| title_full_unstemmed | Improved coordinate attention network for classification of dangerous driving behavior |
| title_short | Improved coordinate attention network for classification of dangerous driving behavior |
| title_sort | improved coordinate attention network for classification of dangerous driving behavior |
| topic | Coordinate attention Dangerous driving Exponential cross entropy |
| url | http://www.sciencedirect.com/science/article/pii/S277318632500009X |
| work_keys_str_mv | AT wenni improvedcoordinateattentionnetworkforclassificationofdangerousdrivingbehavior AT lufengbai improvedcoordinateattentionnetworkforclassificationofdangerousdrivingbehavior |