Learning Spatio-Temporal Representations With a Dual-Stream 3-D Residual Network for Nondriving Activity Recognition
文献类型:期刊论文
作者 | Yang, Lichao1; Shan, Xiaocai3; Lv, Chen2; Brighton, James1; Zhao, Yifan1 |
刊名 | IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS |
出版日期 | 2022-07-01 |
卷号 | 69期号:7页码:7405-7414 |
ISSN号 | 0278-0046 |
关键词 | Three-dimensional displays Vehicles Feature extraction Solid modeling Computational modeling Activity recognition Videos Action recognition automated driving nondriving related task |
DOI | 10.1109/TIE.2021.3099254 |
英文摘要 | Accurate recognition of nondriving activity (NDA) is important for the design of intelligent human machine interface to achieve a smooth and safe control transition in the conditionally automated driving vehicle. However, some characteristics of such activities like limited-extent movement and similar background pose a challenge to the existing 3-D convolutional neural network based action recognition methods. In this article, we propose a dual-stream 3-D residual network, named DS3D residual network (ResNet), to enhance the learning of spatio-temporal representation and improve the activity recognition performance. Specifically, a parallel two-stream structure is introduced to focus on the learning of short-time spatial representation and small-region temporal representation. A two-feed driver behavior monitoring framework is further build to classify four types of NDAs and two types of driving behavior based on the driver's head and hand movement. A novel NDA dataset has been constructed for the evaluation, where the proposed DS3D ResNet achieves 83.35% average accuracy, at least 5% above three selected state-of-the-art methods. Furthermore, this study investigates the spatio-temporal features learned in the hidden layer through the saliency map, which explains the superiority of the proposed model on the selected NDAs. |
资助项目 | Cranfield's EPSRC[EP/R511511/1] |
WOS研究方向 | Automation & Control Systems ; Engineering ; Instruments & Instrumentation |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:000753527500089 |
资助机构 | Cranfield's EPSRC ; Cranfield's EPSRC ; Cranfield's EPSRC ; Cranfield's EPSRC ; Cranfield's EPSRC ; Cranfield's EPSRC ; Cranfield's EPSRC ; Cranfield's EPSRC |
源URL | [http://ir.iggcas.ac.cn/handle/132A11/104927] |
专题 | 地质与地球物理研究所_中国科学院油气资源研究重点实验室 |
通讯作者 | Zhao, Yifan |
作者单位 | 1.Cranfield Univ, Sch Aerosp Transport & Mfg, Cranfield MK43 0AL, Beds, England 2.Nanyang Technol Univ, Sch Elect & Elect Engn, Singapore 639798, Singapore 3.Chinese Acad Sci, Inst Geol & Geophys, Beijing 100029, Peoples R China |
推荐引用方式 GB/T 7714 | Yang, Lichao,Shan, Xiaocai,Lv, Chen,et al. Learning Spatio-Temporal Representations With a Dual-Stream 3-D Residual Network for Nondriving Activity Recognition[J]. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,2022,69(7):7405-7414. |
APA | Yang, Lichao,Shan, Xiaocai,Lv, Chen,Brighton, James,&Zhao, Yifan.(2022).Learning Spatio-Temporal Representations With a Dual-Stream 3-D Residual Network for Nondriving Activity Recognition.IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,69(7),7405-7414. |
MLA | Yang, Lichao,et al."Learning Spatio-Temporal Representations With a Dual-Stream 3-D Residual Network for Nondriving Activity Recognition".IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 69.7(2022):7405-7414. |
入库方式: OAI收割
来源:地质与地球物理研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。