浏览全部资源
扫码关注微信
基于主动学习和Wi-Fi感知的人体识别系统
Human activity recognition system based on active learning and Wi-Fi sensing
- 物联网学报 2022年6卷第1期 页码：44-52
- 作者机构：
1. 中国科学技术大学，安徽 合肥 230026
2. 中国科学技术大学附属第一医院（安徽省立医院），安徽 合肥 230001
- 作者简介：
[ "赵广智（1995− ），男，中国科学技术大学计算机科学与技术学院硕士生，主要研究方向为主动学习、人体行为识别、无线感知等" ]
[ "周志鹏（1994− ），男，中国科学技术大学计算机科学与技术学院博士生，主要研究方向为机器学习、无线感知等" ]
[ "龚伟（1982− ），男，中国科学技术大学教授、博士生导师，主要研究方向为大规模物联网、智能感知、无线网络、分布式计算等" ]
[ "陈绍青（1985− ），男，博士，中国科学技术大学信息科学实验中心讲师，主要研究方向为振动主动控制、自适应控制等" ]
[ "周浩泉（1968− ），男，中国科学技术大学附属第一医院（安徽省立医院）副教授、硕士生导师，主要研究方向医学信息技术等。" ]
- 基金信息：合肥市自主创新政策“借转补”项目;The Independent Innovation Policy Program of Hefei(J2020Y03)
DOI：10.11959/j.issn.2096-3750.2022.00262
中图分类号： TP393纸质出版日期：2022-03-30，
网络出版日期：2022-03，
- 稿件说明：
移动端阅览
赵广智, 周志鹏, 龚伟, 等. 基于主动学习和Wi-Fi感知的人体识别系统[J]. 物联网学报, 2022,6(1):44-52.
GUANGZHI ZHAO, ZHIPENG ZHOU, WEI GONG, et al. Human activity recognition system based on active learning and Wi-Fi sensing. [J]. Chinese journal on internet of things, 2022, 6(1): 44-52.
赵广智, 周志鹏, 龚伟, 等. 基于主动学习和Wi-Fi感知的人体识别系统[J]. 物联网学报, 2022,6(1):44-52. DOI： 10.11959/j.issn.2096-3750.2022.00262.
GUANGZHI ZHAO, ZHIPENG ZHOU, WEI GONG, et al. Human activity recognition system based on active learning and Wi-Fi sensing. [J]. Chinese journal on internet of things, 2022, 6(1): 44-52. DOI： 10.11959/j.issn.2096-3750.2022.00262.
摘要
基于深度学习和Wi-Fi感知的人体行为识别系统已逐步成为主流的研究方向，在近年来得到了长足的发展。然而，现有的系统严重依赖于大量带标记样本以达到良好的识别精度。这导致了大量的人力成本用于标记数据，同时现有系统也难以应用于实际场景。针对该问题，提出一种将主动学习应用于Wi-Fi感知的人体行为识别系统——ALSensing。该系统是第一个将主动学习和Wi-Fi人体行为识别相结合的系统，能够利用有限数量的已标记训练样本构建一个性能良好的行为识别器。利用商用的Wi-Fi设备实现了ALSensing系统，并且使用6个不同场景的实际数据集评估了它的性能。实验结果显示，ALSensing 利用 3.7%的已标记训练样本能够达到 52.83%的识别精度，利用 15%的已标记训练样本能够达到 58.97%的识别精度，而利用现有基于深度学习的人体行为识别系统测量的参考基准在100%的已标记训练样本的情况下达到62.19%的识别精度。可见，ALSensing能够实现与现有基于深度学习的人体行为识别系统接近的识别精度，但是所需要的已标记训练样本的数量大幅度减少。
Abstract
Human activity recognition system based on deep learning and Wi-Fi sensing has gradually become the mainstream research field and has been developed in recent years.However
related systems heavily rely on training with huge labeled samples to reach a high accuracy
which is labor-intensive and unrealistic for many real-world scenarios.To solve this problem
a system that combines active learning with Wi-Fi based human activity recognition—ALSensing was proposed
which was able to train a well-perform classifier with limited labeled samples.ALSensing was implemented with commercial Wi-Fi devices and evaluated in six real environments.The experimental results show that ALSensing achieves 52.83% recognition accuracy using 3.7% of total training samples
58.97% recognition accuracy using 15% of total training samples
while the existing full-supervised system reaches 62.19% recognition accuracy.It demonstrates that ALSensing has a similar performance with baseline but requires much less labeled samples.
关键词
主动学习人体行为识别Wi-Fi
Keywords
active learninghuman activity recognitionWi-Fi
references
刘希文, 陈海明 . 一种基于CSI的人体动作计数与识别方法[J]. 北京邮电大学学报, 2020,43(5): 105-111.
LIU X W, CHEN H M . A human action counting and recognition method based on CSI[J]. Journal of Beijing University of Posts and Telecommunications, 2020,43(5): 105-111.
赵雄, 陈平, 潘晋孝 . 基于骨架关键点的车内异常行为识别方法[J]. 机械与电子, 2021,39(3): 10-15.
ZHAO X, CHEN P, PAN J X . Recognition method of abnormal behavior in car based on skeleton key points[J]. Machinery ＆ Electronics, 2021,39(3): 10-15.
冯仕民, 刘忠育, 俞啸 ,等. 矿山物联网人员情境描述与不安全行为识别[J]. 物联网学报, 2018,2(4): 93-98.
FENG S M, LIU Z Y, YU X ,et al. Workers' context description and unsafe behavior recognition in internet of things for mines[J]. Chinese Journal on Internet of things, 2018,2(4): 93-98.
邓颖, 吴华瑞, 孙想 . 基于机器视觉和穿戴式设备感知的村镇老年人跌倒监测方法[J]. 西南大学学报(自然科学版), 2021,43(11): 186-194.
DENG Y, WU H R, SUN X . Design of a real-time human falling monitoring method for elderly people in villages and towns based on multi-dimensional data analysis[J]. Journal of Southwest University (Natural Science Edition), 2021,43(11): 186-194.
CHEN K, FORBUS K . Action recognition from skeleton data via analogical generalization over qualitative representations[EB]. 2018.
RANJAN J, WHITEHOUSE K . Object hallmarks:identifying object users using wearable wrist sensors[C]// Proceedings of UbiComp '15:Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing. 2015: 51-61.
YUE S C, YANG Y Z, WANG H ,et al. BodyCompass:Monitoring sleep posture with wireless signals[J]. Proceedings of the ACM on Interactive,Mobile,Wearable and Ubiquitous Technologies, 2020,4(2): 1-25.
WANG Y X, WU K S, NI L M . WiFall:device-free fall detection by wireless networks[J]. IEEE Transactions on Mobile Computing, 2017,16(2): 581-594.
ZHAO MM, LI T H, ALSHEIKH M A ,et al. Through-wall human pose estimation using radio signals[C]// Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE Press, 2018: 7356-7365.
FAN L J, LI T H, YUAN Y ,et al. In-home daily-life captioning using radio signals[M]// Computer Vision – ECCV2020. Cham: Springer International Publishing, 2020: 105-123.
WANG K Z, ZHANG D Y, LI Y ,et al. Cost-effective active learning for deep image classification[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2017,27(12): 2591-2600.
KRIZHEVSKY A, SUTSKEVER I, HINTON G E . ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017,60(6): 84-90.
0
浏览量
524
下载量
0
CSCD
- 网络PDF下载
- Meta-XML下载
- BibTeX下载
- Endnote下载
- RefMan 下载
- NoteExpress下载
- NoteFirst下载
关联资源
相关文章
相关作者
相关机构
- 技术支持由北京北大方正电子有限公司提供 京ICP备09064830号-19
京公网安备11010802024621 - 本系统建议在Chrome、 IE9+ 以上版本浏览器阅读本站内容，360浏览器请切换至极速模式
- Cookies帮助我们提供服务并提供个性化体验。使用本网站，即表示您同意我们使用Cookies
- 总访问量：42508 今日访问量：303