Lightweight traffic sign recognition model based on coordinate attention

doi:10.11772/j.issn.1001-9081.2022010100

Journal of Computer Applications ›› 2023, Vol. 43 ›› Issue (2): 608-614.DOI: 10.11772/j.issn.1001-9081.2022010100

• Multimedia computing and computer simulation • Previous Articles

Lightweight traffic sign recognition model based on coordinate attention

Wenju LI, Gan ZHANG, Liu CUI(), Wanghui CHU

School of Computer Science and Information Engineering，Shanghai Institute of Technology，Shanghai 201418，China

Received:2022-01-25 Revised:2022-04-25 Accepted:2022-04-26 Online:2022-05-31 Published:2023-02-10
Contact: Liu CUI
About author:LI Wenju， born in 1964， Ph. D.， professor. His research interests include computer vision， pattern recognition， intelligent detection.
ZHANG Gan， born in 1997， M. S. candidate. His research interests include object detection， traffic sign recognition.
CHU Wanghui， born in 1998， M. S. candidate. Her research interests include graph neural network， 3D object detection.
Supported by:
National Natural Science Foundation of China(61903256)

基于坐标注意力的轻量级交通标志识别模型

李文举, 张干, 崔柳(), 储王慧

上海应用技术大学计算机科学与信息工程学院，上海 201418

通讯作者: 崔柳
作者简介:李文举（1964—），男，辽宁营口人，教授，博士，CCF会员，主要研究方向：计算机视觉、模式识别、智能检测
张干（1997—），男，江苏徐州人，硕士研究生，主要研究方向：目标检测、交通标志识别
储王慧（1998—），女，安徽池州人，硕士研究生，主要研究方向：图神经网络、3D目标检测。
基金资助:
国家自然科学基金资助项目(61903256)

Abstract

Abstract:

For the problems of unbalanced detection speed and recognition accuracy of traffic sign recognition models， and that it is difficult to detect occluded targets and small targets， YOLOv5 （You Only Look Once version 5） model was improved， and a lightweight traffic sign recognition model based on Coordinate Attention （CA） was proposed. Firstly， CA mechanism was integrated into the backbone network to effectively capture the relationships between location information and channels， so as to obtain the regions of interest more accurately and avoid too much computational overhead. Then， cross layer connections were added to the feature fusion network to fuse more feature information without increasing the cost， improve the feature extraction ability of the network and the detection effect of occluded targets. Finally， the improved CIoU （Complete Intersection over Union） function was introduced to calculate the localization loss， thereby alleviating the uneven distribution of sample size in the detection process， and further improving the recognition accuracy of small targets. Applying this model on TT100K （Tsinghua-Tencent 100K） dataset， the recognition accuracy is 91.5%， the recall is 86.64%， which are improved by 20.96% and 11.62% respectively compared with those of the traditional YOLOv5n model， and the frame processing rate is 140.84 FPS （Frames Per Second）. These experimental results fully verify the accuracy and real-time performance of the proposed model for traffic sign detection and recognition in real scenes.

Key words: YOLOv5 (You Only Look Once version 5), traffic sign recognition, Coordinate Attention (CA), feature fusion, loss function

摘要：

针对交通标志识别模型检测速度与识别精度不均衡，以及受遮挡目标和小目标难以检测的问题，对YOLOv5模型进行改进，提出一种基于坐标注意力（CA）的轻量级交通标志识别模型。首先，通过在主干网络中融入CA机制，有效地捕获位置信息和通道之间的关系，从而更准确地获取感兴趣区域，避免过多的计算开销；然后，通过在特征融合网络中加入跨层连接，在不增加成本的情况下融合更多的特征信息，提高网络的特征提取能力，并改善对遮挡目标的检测效果；最后，引入改进的CIoU函数计算定位损失，以缓解检测过程中样本尺寸分布不均衡的现象，并进一步提高对小目标的识别精度。在TT100K数据集上应用所提模型时，识别精度达到了91.5%，召回率达到了86.64%，与传统的YOLOv5n模型相比分别提高了20.96%和11.62%，且帧处理速率达到了140.84 FPS。实验结果比较充分地验证了所提模型在真实场景中对交通标志检测与识别的准确性与实时性。

关键词: YOLOv5, 交通标志识别, 坐标注意力, 特征融合, 损失函数

CLC Number:

TP391.4

Wenju LI, Gan ZHANG, Liu CUI, Wanghui CHU. Lightweight traffic sign recognition model based on coordinate attention[J]. Journal of Computer Applications, 2023, 43(2): 608-614.

李文举, 张干, 崔柳, 储王慧. 基于坐标注意力的轻量级交通标志识别模型[J]. 《计算机应用》唯一官方网站, 2023, 43(2): 608-614.

Figures/Tables 12

References 22

1	HE S H， CHEN L， ZHANG S Y， et al. Automatic recognition of traffic signs based on visual inspection［J］. IEEE Access， 2021， 9： 43253-43261. 10.1109/access.2021.3059052
2	于硕. 交通标志识别技术综述［J］. 科技资讯， 2019， 17（6）： 15-16.
	YU S. Overview of traffic sign recognition technology［J］. Science and Technology Information， 2019， 17（6）： 15-16.
3	FLEYEH H， BISWAS R， DAVAMI E. Traffic sign detection based on AdaBoost color segmentation and SVM classification［C］// Proceedings of the EuroCon 2013. Piscataway： IEEE， 2013： 2005-2010. 10.1109/eurocon.2013.6625255
4	杜影丽，贾永红，韩静敏. 自然场景车载视频道路交通限速标志的检测与识别方法［J］. 测绘地理信息， 2018， 43（2）： 32-34， 37. 10.14188/j.2095-6045.2018018
	DU Y L， JIA Y H， HAN J M. A detection and recognition method for traffic speed limit signs based on vehicle videos［J］. Journal of Geomatics， 2018， 43（2）： 32-34， 37. 10.14188/j.2095-6045.2018018
5	陈名松，吴冉冉，张泽功，等. 基于改进CapsNet的交通标志分类模型［J］. 计算机应用研究， 2020， 37（S2）：367-368， 371.
	CHEN M S， WU R R， ZHANG Z G， et al. Traffic sign classification model based on improved CapsNet［J］. Application Research of Computers， 2020， 37（S2）：367-368， 371.
6	郭璠，张泳祥，唐琎，等. YOLOv3-A：基于注意力机制的交通标志检测网络［J］. 通信学报， 2021， 42（1）：87-99.
	GUO F， ZHANG Y X， TANG J， et al. YOLOv3-A： a traffic sign detection network based on attention mechanism［J］. Journal on Communications， 2021， 42（1）：87-99.
7	JIN Y M， FU Y S， WANG W Q， et al. Multi-feature fusion and enhancement single shot detector for traffic sign recognition［J］. IEEE Access， 2020， 8： 38931-38940. 10.1109/access.2020.2975828
8	WANG C Y， LIAO H Y M， WU Y H， et al. CSPNet： a new backbone that can enhance learning capability of CNN［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Piscataway： IEEE， 2020： 1571-1580. 10.1109/cvprw50498.2020.00203
9	HOU Q B， ZHOU D Q， FENG J S. Coordinate attention for efficient mobile network design［C］// Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2021： 13708-13717. 10.1109/cvpr46437.2021.01350
10	TAN M X， PANG R M， LE Q V. EfficientDet： scalable and efficient object detection［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2020： 10778-10787. 10.1109/cvpr42600.2020.01079
11	LIU S， QI L， QIN H F， et al. Path aggregation network for instance segmentation［C］// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2018： 8759-8768. 10.1109/cvpr.2018.00913
12	ZHENG Z H， WANG P， LIU W， et al. Distance-IoU loss： faster and better learning for bounding box regression［C］// Proceedings of the 34th AAAI Conference on Artificial Intelligence. Palo Alto， CA： AAAI Press， 2020： 12993-13000. 10.1609/aaai.v34i07.6999
13	ZHU Z， LIANG D， ZHANG S H， et al. Traffic-sign detection and classification in the wild［C］// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2016： 2110-2118. 10.1109/cvpr.2016.232
14	XU K， BA J， KIROS R， et al. Show， attend and tell： neural image caption generation with visual attention［C］// Proceedings of the 32nd International Conference on Machine Learning. New York： JMLR.org， 2015： 2048-2057. 10.1109/cvpr.2015.7298935
15	HU J， SHEN L， SUN G. Squeeze-and-excitation networks［C］// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2018： 7132-7141. 10.1109/cvpr.2018.00745
16	WOO S， PARK J， LEE J Y， et al. CBAM： convolutional block attention module［C］// Proceedings of the 2018 European Conference on Computer Vision， LNCS 11211. Cham： Springer， 2018： 3-19.
17	LIN T Y， DOLLÁR P， GIRSHICK R， et al. Feature pyramid networks for object detection［C］// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2017： 936-944. 10.1109/cvpr.2017.106
18	HE K M， ZHANG X Y， REN S Q， et al. Spatial pyramid pooling in deep convolutional networks for visual recognition［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2015， 37（9）： 1904-1916. 10.1109/tpami.2015.2389824
19	JIANG B R， LUO R X， MAO J Y， et al. Acquisition of localization confidence for accurate object detection［C］// Proceedings of the 2018 European Conference on Computer Vision， LNCS 11218. Cham： Springer， 2018： 816-832.
20	REZATOFIGHI H， TSOI N， GWAK J， et al. Generalized intersection over union： a metric and a loss for bounding box regression［C］// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2019： 658-666. 10.1109/cvpr.2019.00075
21	龚祎垄，吴勇，陈铭峥. 针对TT100K交通标志数据集的扩增策略［J］. 福建电脑， 2019， 35（11）：70-71.
	GONG Y L， WU Y， CHEN M Z. An enlargement strategy for TT100K traffic sign data set［J］. Journal of Fujian Computer， 2019， 35（11）：70-71.
22	张干，李文举，张耀星. 基于改进的YOLOv5算法的交通标志识别［C］// 21全国仿真技术学术会议论文集. 北京：计算机仿真杂志社， 2021：182-185， 249. 10.1109/iccnea50255.2020.00021
	ZHANG G， LI W J， ZHANG Y X. Traffic sign recognition based on improved YOLOv5 algorithm［C］// Proceedings of the 2021 China Simulation Technology Conference. Beijing： Periodical Office of Computer Simulation， 2021：182-185， 249. 10.1109/iccnea50255.2020.00021

模型	精度	召回率	帧处理速率/FPS	参数量/10⁶	模型大小/MB
YOLOv5n	0.756 4	0.776 2	333.33	1.820	3.85
YOLOv5s	0.820 3	0.859 0	250.00	7.140	14.00
YOLOv5m	0.825 6	0.880 1	128.20	21.040	40.60
YOLOv5l	0.830 0	0.890 0	75.75	46.370	80.90
YOLOv5x	0.853 0	0.913 0	44.64	86.510	165.00
YOLOv5n （1 280×1 280）	0.868 1	0.850 0	156.25	1.824	4.44

模型	精度	召回率	帧处理速率/FPS	参数量/10⁶	模型大小/MB
YOLOv5n	0.756 4	0.776 2	333.33	1.820	3.85
YOLOv5s	0.820 3	0.859 0	250.00	7.140	14.00
YOLOv5m	0.825 6	0.880 1	128.20	21.040	40.60
YOLOv5l	0.830 0	0.890 0	75.75	46.370	80.90
YOLOv5x	0.853 0	0.913 0	44.64	86.510	165.00
YOLOv5n （1 280×1 280）	0.868 1	0.850 0	156.25	1.824	4.44

模型	精度	召回率	帧处理速率/FPS	参数量/10⁶	模型大小/MB
YOLOv5n-P6	0.815 7	0.820 0	263.15	3.172	6.51
YOLOv5s-P6	0.840 1	0.835 1	208.30	12.480	24.30
YOLOv5m-P6	0.865 4	0.854 9	111.11	35.530	68.30
YOLOv5l-P6	0.872 3	0.871 0	68.02	76.500	146.00
YOLOv5x-P6	0.8762	0.8786	40.32	140.450	268.00
YOLOv5n-P6 （1 280×1 280）	0.875 0	0.852 0	149.25	3.180	7.13

模型	精度	召回率	帧处理速率/FPS	参数量/10⁶	模型大小/MB
YOLOv5n-P6	0.815 7	0.820 0	263.15	3.172	6.51
YOLOv5s-P6	0.840 1	0.835 1	208.30	12.480	24.30
YOLOv5m-P6	0.865 4	0.854 9	111.11	35.530	68.30
YOLOv5l-P6	0.872 3	0.871 0	68.02	76.500	146.00
YOLOv5x-P6	0.8762	0.8786	40.32	140.450	268.00
YOLOv5n-P6 （1 280×1 280）	0.875 0	0.852 0	149.25	3.180	7.13

模型	精度	召回率	帧处理速率/FPS	参数量/10⁶	模型大小/MB
YOLOv5n-P6+ 坐标注意力	0.847 0	0.833 5	232.56	3.210	6.6
YOLOv5s-P6+ 坐标注意力	0.851 9	0.836 3	217.39	12.620	24.5
YOLOv5m-P6+ 坐标注意力	0.835 2	0.815 8	107.53	35.930	69.1
YOLOv5l-P6+ 坐标注意力	0.826 0	0.811 2	64.52	77.390	148.0
YOLOv5x-P6+ 坐标注意力	0.823 2	0.800 3	36.63	142.140	272.0
YOLOv5n-P6+ 坐标注意力（1 280×1 280）	0.8987	0.8535	144.93	3.216	7.22

Lightweight traffic sign recognition model based on coordinate attention

基于坐标注意力的轻量级交通标志识别模型

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 12

References 22

Related Articles 15

Recommended Articles

Metrics

模型	精度	召回率	帧处理速率/FPS	参数量/10⁶	模型大小/MB
YOLOv3	0.820 0	0.831 4	36.76	58.74	237.00
RetinaNet-NeXt	0.874 5	0.790 0
YOLOv3-A	0.885 0	0.922 0	1.25
YOLOv4	0.869 0	0.889 0	35.84		244.00
YOLOX-Nano	0.613 0		349.65	2.25	17.40
Faster R-CNN	0.715 0	0.765 0	11.11		159.54
FA-SSD	0.802 0		13.60
YOLOv5n-P6+ 坐标注意力+ 跨层连接（1 280×1 280）	0.910 5	0.855 0	140.84	3.26	7.32
本文模型	0.915 0	0.866 4	140.84	3.26	7.32

[1]	Ping WANG, Nan CHEN, Lei LU. Fall detection algorithm based on scene prior and attention guidance [J]. Journal of Computer Applications, 2023, 43(2): 529-535.
[2]	Gang CHEN, Yongwei LIAO, Zhenguo YANG, Wenying LIU. Image inpainting algorithm of multi-scale generative adversarial network based on multi-feature fusion [J]. Journal of Computer Applications, 2023, 43(2): 536-544.
[3]	Kewen LI, Jiantao YANG, Zongchao HUANG. Improved YOLOv3 target detection based on boundary limit point features [J]. Journal of Computer Applications, 2023, 43(1): 81-87.
[4]	Honggang YANG, Jiejie CHEN, Mengfei XU. Bilinear involution neural network for image classification of fundus diseases [J]. Journal of Computer Applications, 2023, 43(1): 259-264.
[5]	Yuchao LYU, Xi JIANG, Yinghao XU, Xijun ZHU. Improved brachial plexus nerve segmentation method based on multi-scale feature fusion [J]. Journal of Computer Applications, 2023, 43(1): 273-279.
[6]	Zanxia QIANG, Xianfu BAO. Residual attention deraining network based on convolutional long short-term memory [J]. Journal of Computer Applications, 2022, 42(9): 2858-2864.
[7]	Jinghan YIN, Shaojun QU, Zekai YAO, Xuanye HU, Xiaoyu QIN, Pujing HUA. Traffic sign recognition model in haze weather based on YOLOv5 [J]. Journal of Computer Applications, 2022, 42(9): 2876-2884.
[8]	Yuyu MENG, Jing GUO. Link prediction algorithm based on information entropy improved PCA model [J]. Journal of Computer Applications, 2022, 42(9): 2823-2829.
[9]	Yuhang WANG, Yongxia ZHOU, Liangwu WU. Pooling algorithm based on Gaussian function [J]. Journal of Computer Applications, 2022, 42(9): 2800-2806.
[10]	Ning DONG, Xiaorong CHENG, Mingquan ZHANG. Intrusion detection system with dynamic weight loss function based on internet of things platform [J]. Journal of Computer Applications, 2022, 42(7): 2118-2124.
[11]	Dawei ZHANG, Xuchong LIU, Wei ZHOU, Zhuhui CHEN, Yao YU. Real-time traffic sign detection algorithm based on improved YOLOv3 [J]. Journal of Computer Applications, 2022, 42(7): 2219-2226.
[12]	Ruijie YANG, Guilin ZHENG. Face liveness detection based on InceptionV3 and feature fusion [J]. Journal of Computer Applications, 2022, 42(7): 2037-2042.
[13]	Jin ZHANG, Peiqi QU, Cheng SUN, Meng LUO. Safety helmet wearing detection algorithm based on improved YOLOv5 [J]. Journal of Computer Applications, 2022, 42(4): 1292-1300.
[14]	Zhihua LIU, Wenjie CHEN, Aibin CHEN. Homologous spectrogram feature fusion with self-attention mechanism for bird sound classification [J]. Journal of Computer Applications, 2022, 42(4): 1260-1268.
[15]	Xinrong HU, Junyu ZHANG, Tao PENG, Junping LIU, Ruhan HE, Kai HE. Cascaded cross-domain feature fusion for virtual try-on [J]. Journal of Computer Applications, 2022, 42(4): 1269-1274.