[1]马迎杰,王佳斌,郑力新,等.深度可分离卷积网络的驾驶状态识别算法[J].华侨大学学报(自然科学版),2021,42(2):259-267.[doi:10.11830/ISSN.1000-5013.202001010]
　MA Yingjie,WANG Jiabin,ZHENG Lixin,et al.Driving State Recognition Algorithm Based on Deep Separable Convolutional Network[J].Journal of Huaqiao University(Natural Science),2021,42(2):259-267.[doi:10.11830/ISSN.1000-5013.202001010]
点击复制

深度可分离卷积网络的驾驶状态识别算法()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第42卷
期数:
2021年第2期
页码:
259-267
栏目:
出版日期:
2021-03-20

文章信息/Info

Title:
Driving State Recognition Algorithm Based on Deep Separable Convolutional Network
文章编号:
1000-5013(2021)02-0259-09
作者:
马迎杰;  王佳斌;  郑力新;  朱新龙
华侨大学 工学院, 福建 泉州 362021
Author(s):
MA Yingjie;  WANG Jiabin;  ZHENG Lixin;  ZHU Xinlong
College of Engineering, Huaqiao University, Quanzhou 362021, China
关键词:
驾驶状态;  状态特征检测;  深度学习;  深度卷积;  逐点卷积
Keywords:
driving state;  state feature detection;  deep learning;  depthwise convolution;  pointwise convolution
分类号:
TP391
DOI:
10.11830/ISSN.1000-5013.202001010
文献标志码:
A
摘要:
针对嵌入式设备内存小及多分类准确率低等导致驾驶员检测问题,提出经过深度可分离卷积网络改进而成的,快速下采样网络(fast downsampling network,MF-Net)驾驶状态识别系统.即将快速下采样策略应用于深度可分离卷积网络,在12层内执行32倍下采样,以有效降低计算成本、增加信息容量并实现性能改进.实验结果表明:与VGG(visual geometry group)和ResNet 50等其他卷积神经网络(CNN)模型相比,所提出的MF-Net模型深度可分离卷积大大减少参数量,快速下采样方案的运用增加了网络的信息容量,不仅模型较小且在驾驶员状态分类方面能够表现出更好的性能.同时,信息容量的增加可以对更多信息进行编码,加深对图像内容的理解,有利于之后的嵌入式系统移植.
Abstract:
Aiming at the problem of driver detection caused by the small memory of embedded devices and the low accuracy of multi-classification, a fast fast down sampling network(MF-Net)driving state recognition system improved by deep separable convolutional network is proposed, the key idea is applying a fast downsampling strategy to deep separable convolutional networks, which performs 32-fold downsampling within 12 layers to reduce computational cost seffectively, increase information capacity, and achieve performance improvements. The experimental results show that: compared with other convolutional neural network(CNN)models such as VGG(visual geometry group)and ResNet 50, MF-Net model has deep separable convolutions that reduce the amount of parameters greatly and the application of fast down sampling increase the information capacity of the network, the model not only smaller but also show better performance in the classification of driver status, at the same time, the increase in information capacity can encode more information and deepen the understanding of the image content, which is beneficial to transplantation of embedded systemsin the future.

参考文献/References:

[1] TAMAS V,MATIES V,NNEBE S E,et al.Real-time distracted drivers detection using deep learning[J].American Journal of Artificial Intelligence,2019,3(1):1-8.
[2] 吴绍斌,高利,王刘安.基于脑电信号的驾驶疲劳检测研究[J].北京理工大学学报,2009,29(12):1072-1075.DOI:10.15918/j.tbi t1001-0645.2009.12.013.
[3] JAP B T,LAL S,FISCHER P,et al.Using EEG spectral components to assess algorithms for detecting fatigue[J].Expert Systems with Applications,2009,36(2):2352-2359.DOI:10.1016/j.eswa.2007.12.043.
[4] LI Mingai,ZHANG Cheng,YANG Jinfu.An EEG-based method for detecting drowsy driving state[C]//Seventh International Conference on Fuzzy Systems and Knowledge Discovery.Yantai: IEEE Press,2010:2164-2167.DOI:10.1109/FSKD.2010.5569757.
[5] 王玉海,宋健,李兴坤.基于模糊推理的驾驶员意图识别研究[J].公路交通科技,2005,22(12):116-121,142.DOI:10.3969/j.issn.1002-0268.2005.12.030.
[6] ZHAO C H,ZHANG Bailing,HE Jie,et al.Recognition of driving postures by contourlet transform and random forests[J].IET Intelligent Transport Systems,2012,6(2):161-168.DOI:10.1049/iet-its.2011.0116
[7] BERRI R A,SILVA A G,PARPINELLI R S,et al.A pattern recognition system for detecting use of mobile phones while driving[C]//International Conference on Computer Vision Theory and Applications(VISAPP).Lisbon: IEEE Press,2014:411-418.DOI:10.5220/0004684504110418.
[8] CRAYE C,KARRAY F.Driver distraction detection and recognition using RGB-D sensor[J/OL].eprint arXiv: 1502.00250,2015:1-11.[2015-02-01] .http://de.arxiv.org/pdf/1502.00250.
[9] LE T H N,ZHENG Yutong,ZHU Chenchen,et al.Multiple scale faster-rcnn approach to driver’s cell-phone usage and hands on steering wheel detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops.Las Vegas: IEEE Press,2016:46-53.DOI:10.1109/CVPRW.2016.13.
[10] RUIZ N,CHONG E,REHG J M.Fine-grained head pose estimation without keypoints[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops.Salt Lake City: IEEE Press,2018:2074-2083.DOI:10.1109/CVPRW.2018.00281.
[11] SIMARD P Y,STEINKRAUS D,PLATT J C.Best practices for convolutional neural networks applied to visual document analysis[C]//Seventh International Conference on Document Analysis and Recognition.Edinburgh: IEEE Press,2003.DOI:10.1109/ICDAR.2003.1227801.
[12] LIN Ming,CHEN Qiang,YAN Shuicheng.Network in network[EB/OL].(2013-12-16)[2014-03-04] .https://arxiv.org/abs/1312.4400.
[13] CHOLLET F.Xception: Deep learning with depthwise separable convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Puerto Rico: IEEE Press,2017:1251-1258.DOI:10.1109/CVPR.2017.195.
[14] BAHETI B,GAJRE S,TALBAR S.Detection of distracted driver using convolutional neural network[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops.Salt Lake City: IEEE Press,2018.DOI:10.1109/CVPRW.2018.00150.
[15] SANG Haifeng,WANG Chuanzheng,HE Dakuo,et al.Multi-information flow CNN and attribute-aided reranking for person reidentification[J/OL].Computational Intelligence and Neuroscience,2019:1-12.http://downloads.hindawi.com/journals/cin/2019/7028107.pdf.DOI:10.1155/2019/7028107.
[16] LI Xiaoguang,LAM K M,QIU G P,et al.Examplebased image super-resolution with class-speciflc predictors[J].Journal of Visual Communication and Image Representation,2009,20(5):312-322.DOI:10.1016/j.jvcir.2009.03.008.
[17] NIE Weizhi,WANG Kun,WANG Hongtao,et al.The assessment of 3D model representation for retrieval with CNN-RNN networks[J].Multimedia Tools and Applications,2019,78(1):16979-16994.DOI:10.1007/s11042-018-7102-2.
[18] JIJI C V,CHAUDHURI S.Single-frame image super-resolution through contourlet learning[J/OL].EURASIP Journal on Advances in Signal Processing,2006.https://asp-eurasipjournals.springeropen.com/track/pdf/10.1155/ASP/2006/73767.DOI:10.1155/asp/2006/73767.
[19] YILDIRIM O,BALOGLU U B,TAN R S,et al.A new approach for arrhythmia classification using deep coded features and LSTM networks[J].Computer Methods and Programs in Biomedicine,2019,176:121-133.DOI:10.1016/j.cmpb.2019.05.004.
[20] 李键红,吴亚榕,吕巨建.基于自相似性与多任务高斯过程回归的单帧图像超分辨率重建[J].光学精密工程,2018,26(11):2814-2826.DOI:10.3788/OPE.20182611.2814.
[21] RAJEEV R,SAMTH J A,KARTHIKEYAN N K.An intelligent recurrent neural network with long short-term memory(LSTM)BASED batch normalization for medical image denoising[J/OL].Journal of Medical Systems,2019,43(8):[2019-06-15] .DOI:10.1007/s10916-019-1371-9.
[22] SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[EB/OL].arXiv:1409.1556,2015.(2014-09-04)[2015-04-10] .http://www.arxiv.org/pdf/1409.1556.pdf.
[23] SHI Wenzhe,CABALLERO J,HUSZáR F,et al.Real-time single image and video super-resolution using an efficient subpixel convolutional neural network[C]//IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Las Vegas: IEEE Press,2016.DOI:10.1109/CVPR.2016.207.
[24] OTSU N.A threshold selection method from gray-level histograms[J].IEEE Transactions on Systems,Man and Cybernetics,1979,9(1):62-66.DOI:10.1109/TSMC.1979.4310076
[25] XU Xiaohong,WU Zhihui,CHEN Yu,et al.Plant root spatial distribution measurements based on the hough transformation[J].Neurocomputing,2014,145:209-220.DOI:10.1016/j.neucom.2014.05.041.

备注/Memo

备注/Memo:
收稿日期: 2020-01-09
通信作者: 王佳斌(1974-),男,副教授,主要从事物联网、云计算、大数据和智能仪器的研究.E-mail:fatwang@hqu.edu.cn.
基金项目: 国家自然科学基金青年科学基金资助项目(61505059); 福建省厦门市科技局产学研协同创新资助项目(3502Z20173046); 华侨大学研究生科研创新能力培育计划资助项目(17013084002)
更新日期/Last Update: 2021-03-20