计算机科学 ›› 2022, Vol. 49 ›› Issue (5): 78-83.doi: 10.11896/jsjkx.210400024

• 计算机图形学&多媒体* 上一篇    下一篇

基于SVM的类别增量人体活动识别方法

邢云冰1, 龙广玉1,2, 胡春雨3, 忽丽莎4   

  1. 1 中国科学院计算技术研究所 北京100190
    2 湘潭大学计算机学院·网络空间安全学院 湖南 湘潭411105
    3 齐鲁工业大学计算机科学与技术学院 济南250353
    4 河北经贸大学 信息技术学院 石家庄050061
  • 收稿日期:2021-04-01 修回日期:2021-10-18 出版日期:2022-05-15 发布日期:2022-05-06
  • 通讯作者: 邢云冰(xingyunbing@ict.ac.cn)
  • 基金资助:
    国家重点研发计划(2018YFC2002603);国家自然科学基金(62002187);河北省高等学校科学技术研究资助项目(QN2018116)

Human Activity Recognition Method Based on Class Increment SVM

XING Yun-bing1, LONG Guang-yu1,2, HU Chun-yu3, HU Li-sha4   

  1. 1 Institute of Computing Technology,Chinese Academy of Sciences,Beijing 100190,China
    2 School of Computer Science & School of Cyberspace Science,Xiangtan University,Xiangtan,Hunan 411105,China
    3 Schoolof Computer Scienceand Technology,Qilu University of Technology,Jinan 250353,China
    4 School of Information Technology,Hebei University of Economics and Business,Shijiazhuang 050061,China
  • Received:2021-04-01 Revised:2021-10-18 Online:2022-05-15 Published:2022-05-06
  • About author:XING Yun-bing,born in 1982,master,senior engineer.His main research interests include sign language interaction,pervasive computing and health surveillance.
  • Supported by:
    National Key Research and Development Program of China(2018YFC2002603),National Natural Science Foundation of China(62002187) and Science and Technology Project of Hebei Education Department(QN2018116).

摘要: 基于人体活动识别(Human Activity Recognition,HAR)的健康监护是发现健康异常的一种重要手段。然而,在日常活动识别中,很难提前获取包含所有可能活动类别的训练样本。当预测阶段出现新增类别时,传统的支持向量机(Support Vector Machine,SVM)会将其错误地分类为已知类别。一个鲁棒的分类器应该能够分辨出新增类别,以便后续区别于已知类别并对其进行处理。文中提出一种基于SVM的类别增量人体活动识别方法,引入超球面的思想,既能高精度地识别已知活动类别,又能检测出新增类别。通过训练得到的多个超球面将整个特征空间进行划分,使分类器具有对新增活动类别的检测能力。实验结果表明,与传统多分类SVM方法相比,该方法能够在不显著降低已知类别分类效果的前提下实现对新增类别的检测,从而提高分类器在开放环境下的人体活动识别能力。

关键词: 超球面, 聚类可分, 类别增量, 人体活动识别, 支持向量机

Abstract: Health monitoring based on human activity recognition (HAR) is an important means to discover health abnormalities.However,in daily activity recognition,it is difficult to obtain training samples containing all possible activity categories in advance.When new categories appear in the prediction stage,the traditional support vector machine (SVM) will incorrectly classify them as known category.A robust classifier should be able to distinguish the newly added categories so that they can be processed differently from the known categories.This paper proposes a human activity recognition method based on class increment SVM,and the idea of hypersphere is introduced,which can not only identify known activity categories with high accuracy,but also detect new categories.The multiple hyperspheres obtained through training divide the entire feature space,so that the classifier has the ability to detect newly added activity categories.The experimental results show that compared with the traditional multi-class SVM method,our method can realize the detection of new categories without significantly reducing the classification effect of known categories,thereby improving the classifier's ability to recognize human activity in an open environment.

Key words: Class increment, Clustering separability, Human activity recognition, Hyperspheres, Support vector machine

中图分类号: 

  • TP391.41
[1]LARA O D,LABABRADOR M A.A Survey on Human Acti-vity Recognition Using Wearable Sensors[J].IEEE Communications Surveys & Tutorials,2012,15(3):1192-1209.
[2]ZHANG C X,ZHAO C L,CHEN C,et al.Review of Human Activity Recognition Based on Mobile Phone Sensors[J].Computer Science,2020,47(10):1-8.
[3]DAVIDE A,ALESSANDRO G,LUCA O,et al.A Public Do-main Dataset for Human Activity Recognition Using Smartphones[C]//21th European Symposium on Artificial Neural Networks Computational Intelligence and Machine Learning(ESANN 2013).Bruges,Belgium,2013:24-26.
[4]WANG Z M,WANG K,HE Y.Human Motion Activity Recognition Model Based on Multi-classifier Fusion[J].Computer Science,2016,43(12):297-301.
[5]YANG X,CHEN Y,YU H,et al.Instance-Wise Dynamic Sensor Selection for Human Activity Recognition[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020:1104-1111.
[6]YANG W B,YANG H C,LU C,et al.Gesture RecognitionBased on Skin Color Features and Convolutional Neural Network[J].Journal of Chongqing Technology and Business University(Natural Science Edition),2018,35(4):75-81.
[7]SONG X R,ZHANG X Q,ZHAN Z,et al.Multi-Sensor DataFusion for Complex Human Activity Recognition[J].Journal of Tsinghua University(Science and Technology),2020,60(10):814-821.
[8]DONG L H,LIU Q,CHEN H M,et al.A Lightweight Real-Time Motion Recognition Algorithm Based on Time Window[J].Journal of Computer Research and Development,2017,54(12):2731-2740.
[9]SHAAFI A,SALEM O,MEHAOUA A.Improving Human Activity Recognition Algorithms Using Wireless Body Sensors and SVM[C]//2020 International Wireless Communications and Mobile Computing (IWCMC).IEEE,2020.
[10]YANG J B,NGUYEN M N,SAN P P,et al.Deep Convolutional Neural Networks on Multichannel Time Series for Human Activity Recognition[C]//International Conference on Artificial Intelligence.2015:3995-4001.
[11]ROKNI S A,NOUROLLAHI M,GHASEMZADEH H.Perso-nalized Human Activity Recognition Using Convolutional Neural Networks[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2018.
[12]DENG S Z,WANG B T,YANG C G et al.Convolutional Neural Networks for Human Activity Recognition Using Multi-location Wearable Sensors[J].Journal of Software,2019,30(3):718-737.
[13]LI B,CUI W,WANG W,et al.Two-Stream Convolution Augmented Transformer for Human Activity Recognition[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2021.
[14]GENG C,HUANG S J,CHEN S.Recent Advances in Open Set Recognition:A Survey[J].arXiv:1811.08581,2020.
[15]SCHEIRER W J,JAIN L P,BOULT T E.Probability Models for Open Set Recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(11):2317-2324.
[16]BUSTO P P,GALL J.Open Set Domain Adaptation[C]//IEEE International Conference on Computer Vision.IEEE Computer Society,2017:754-763.
[17]VARETOR,SILVA S,COSTA F,et al.Towards Open-Set Face Recognition Using Hashing Functions[C]//2018 IEEE International Joint Conference on Biometrics (IJCB).IEEE,2018.
[18]ANGUITA D,GHIO A,ONETO L,et al.A Public Domain Dataset for Human Activity Recognition Using Smartphones[C]//Proceedings of the 21th European Symposium on Artificial Neural Networks,Computational Intelligence and Machine Learning(ESANN).2013:437-442.
[19]DUA D,GRAFF C.UCI Machine Learning Repository[EB/OL].https://archive.ics.uci.edu/ml/datasets/Activity+recognition+using+wearable+physiological+measurements.
[20]JIANG X,CHEN Y,LIU J,et al.AIR:Recognizing Activity Through IR-Based Distance Sensing on Feet[C]//Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing:Adjunct.ACM,2016:97-100.
[1] 单晓英, 任迎春.
基于改进麻雀搜索优化支持向量机的渔船捕捞方式识别
Fishing Type Identification of Marine Fishing Vessels Based on Support Vector Machine Optimized by Improved Sparrow Search Algorithm
计算机科学, 2022, 49(6A): 211-216. https://doi.org/10.11896/jsjkx.220300216
[2] 陈景年.
一种适于多分类问题的支持向量机加速方法
Acceleration of SVM for Multi-class Classification
计算机科学, 2022, 49(6A): 297-300. https://doi.org/10.11896/jsjkx.210400149
[3] 侯夏晔, 陈海燕, 张兵, 袁立罡, 贾亦真.
一种基于支持向量机的主动度量学习算法
Active Metric Learning Based on Support Vector Machines
计算机科学, 2022, 49(6A): 113-118. https://doi.org/10.11896/jsjkx.210500034
[4] 武玉坤, 李伟, 倪敏雅, 许志骋.
单类支持向量机融合深度自编码器的异常检测模型
Anomaly Detection Model Based on One-class Support Vector Machine Fused Deep Auto-encoder
计算机科学, 2022, 49(3): 144-151. https://doi.org/10.11896/jsjkx.210100142
[5] 侯春萍, 赵春月, 王致芃.
基于自反馈最优子类挖掘的视频异常检测算法
Video Abnormal Event Detection Algorithm Based on Self-feedback Optimal Subclass Mining
计算机科学, 2021, 48(7): 199-205. https://doi.org/10.11896/jsjkx.200800146
[6] 郭福民, 张华, 胡瑢华, 宋岩.
一种基于表面肌电信号的腕部肌力估计方法研究
Study on Method for Estimating Wrist Muscle Force Based on Surface EMG Signals
计算机科学, 2021, 48(6A): 317-320. https://doi.org/10.11896/jsjkx.200600021
[7] 卓雅倩, 欧博.
噪声环境下的人脸防伪识别算法研究
Face Anti-spoofing Algorithm for Noisy Environment
计算机科学, 2021, 48(6A): 443-447. https://doi.org/10.11896/jsjkx.200900207
[8] 雷剑梅, 曾令秋, 牟洁, 陈立东, 王淙, 柴勇.
基于整车EMC标准测试和机器学习的反向诊断方法
Reverse Diagnostic Method Based on Vehicle EMC Standard Test and Machine Learning
计算机科学, 2021, 48(6): 190-195. https://doi.org/10.11896/jsjkx.200700204
[9] 王友卫, 朱晨, 朱建明, 李洋, 凤丽洲, 刘江淳.
基于用户兴趣词典和LSTM的个性化情感分类方法
User Interest Dictionary and LSTM Based Method for Personalized Emotion Classification
计算机科学, 2021, 48(11A): 251-257. https://doi.org/10.11896/jsjkx.201200202
[10] 曹素娥, 杨泽民.
基于聚类分析算法和优化支持向量机的无线网络流量预测
Prediction of Wireless Network Traffic Based on Clustering Analysis and Optimized Support Vector Machine
计算机科学, 2020, 47(8): 319-322. https://doi.org/10.11896/jsjkx.190800075
[11] 徐翔燕, 侯瑞环.
基于GM(1,1)-SVM组合模型的中长期人口预测研究
Medium and Long-term Population Prediction Based on GM(1,1)-SVM Combination Model
计算机科学, 2020, 47(6A): 485-487. https://doi.org/10.11896/JsJkx.190900168
[12] 马创, 吕孝飞, 梁炎明.
基于GA-SVM的农产品质量分类
Agricultural Product Quality Classification Based on GA-SVM
计算机科学, 2020, 47(6A): 517-520. https://doi.org/10.11896/JsJkx.190900184
[13] 宋岩, 胡瑢华, 郭福民, 袁新亮, 熊睿洋.
基于sEMG的改进SVM+BP肌力预测分层算法
Improved SVM+BP Algorithm for Muscle Force Prediction Based on sEMG
计算机科学, 2020, 47(6A): 75-78. https://doi.org/10.11896/JsJkx.190900143
[14] 方梦琳, 唐文兵, 黄鸿云, 丁佐华.
基于模糊信息分解与控制规则的移动机器人沿墙导航
Wall-following Navigation of Mobile Robot Based on Fuzzy-based Information Decomposition and Control Rules
计算机科学, 2020, 47(6A): 79-83. https://doi.org/10.11896/JsJkx.191000158
[15] 潘恒, 李景峰, 马君虎.
可抵御内部威胁的角色动态调整算法
Role Dynamic Adjustment Algorithm for Resisting Insider Threat
计算机科学, 2020, 47(5): 313-318. https://doi.org/10.11896/jsjkx.190800051
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!