Abstract
This paper presents an inferring and training architecture for long-term and continuous daily activity monitoring using a wearable body sensor network. Energy efficiency and system adaptivity to wearers are two of the most important requirements of a body sensor network. This paper discusses a two-layered hidden Markov model (HMM) architecture for in-network data processing to achieve energy efficiency and model individualization. The bottom-layer HMM is used to process sensory data locally at each wireless sensor node to significantly reduce data transmissions. The top-layer HMM is utilized to find the activity sequence from the result of the local processing. This approach is energy efficient in that only the results of the decoding procedure in each node need to be transmitted rather than raw sensing data. Therefore, the volume of data are significantly reduced. When the algorithm is applied in online monitoring systems, the results of local processing are transmitted only upon hidden state changes. The top-layer processing uses “old data” of one sensor node when it does not receive a “new” result sequence of the local processing from that sensor node. The adaption to various wearers is also discussed, and the robustness of this classification system is depicted. Experiments of 19 activity sequences to be classified are taken by 5 subjects to evaluate the performance of this system.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Apiletti D, Baralis E, Cerquitelli T (2010) Energy-saving models for wireless sensor networks. Knowl Inf Syst 1–30
Andreao RV, Dorizzi B, Boudy J (2006) Ecg signal analysis through hidden markov models. IEEE Trans Biomedical Eng 53(8): 1541–1549
Burchfield TR, Venkatesan S (2007) Accelerometer-based human abnormal movement detection in wireless sesor networks. The 1st international workshop on systems and networking support for healthcare and assisted living environments
Bao L, Intille SS (2004) Activity recognition from user-annotated acceleration data. In: Proceedings of the second international conference on pervasive computing
Baum LE, Petrie T, Soules G et al (1970) A maximization technique occurring in the statistical analysis of probabilistic functions of markov chains. Ann Math Stat 41(1): 164–171
Choudhury T, Philipose M, Wyatt D et al (2006) Towards activity databases: using sensors and statistical models to summarize peoples lives. IEEE Data Eng Bull 29(1):49–58
Chen J, Kwong K, Chang D et al (2005) Wearable sensors for reliable fall detection. In: Proceedings of the 27th annual conference on engineering in medicine and biology
He J, Li H, Tan J (2007) Real-time daily activity classification with wireless sensor networks using hidden markov model. The 29th annual international conference of the IEEE engineering in medicine and biology society, pp 3192–3195
KarantonisDM Narayanan M, Mathie M et al (2006) Implementation of a real-time human movement classifier using a triaxial accelerometer for ambulatory monitoring. IEEE Trans Inf Technol Biomedicine 10(1): 156–167
Li CJ, Khan L, Prabhakaran B (2006) Real-time classification of variable length multi-attribute motions. Knowl Inf Syst 10(2): 163–183
Luo HZ, Fan JP, Lin XD et al (2009) A distributed approach to enabling privacy-preserving model-based classifier training. Knowl Inf Syst 20: 157–185
Mathie MJ, Coster ACF, Lovell NH et al (2003) Detection of daily physical activities using a triaxial accelerometer. Med Biological Eng Comput 41: 296–301
Mathie MJ, Celler BG, Lovell NH et al (2004) Classification of basic daily movements using a triaxial accelerometer. Med Biological Eng Comput 42(5): 679–687
Oliver N, Garg A, Horvitz E (2004) Layered representations for learning and inferring office activity from multiple sensory channels. Comput Vis Image Underst 96: 163–180
Rabiner LR (1989) A tutorial on hidden markov models and selected applications in speech recognition. Proc IEEE 77(2): 257–286
Renevey P, Vetter R, Celka P et al (2002) Activity classification using hmm for improvement of wrist located pulse detection. The 16th international EURASIP conference
Sung M, Marci C, Pentland A (2005) Wearable feedback systems for rehabilitation. J NeuroEngineering and Rehabil 2:17
Salmenkivi M, Mannila H (2005) Using markov chain monte carlo and dynamic programming for event sequence data. Knowl Inf Syst 7: 267–288
Vainer I, Kraus S, Kaminka GA, Slovin H (2010) Obtaining scalable and accurate classification in large-scale spatio-temporal domains. Knowl Inf Syst 1–38
Zhang K, Werner P, Sun M et al (2003) Measurement of human daily physical activity. Obes Res 11(1): 33–40
Author information
Authors and Affiliations
Corresponding author
Additional information
This work was supported in part by the National Science Foundation under Grant ECS #0528967, CNS #0720781, Army Research Laboratory under contract ARL W911NF-06-2-0029 and CERDEC W15P7T-06-P228, Air Force Research Laboratory under a DURIP project and the 863 Program of China under Grant 2009AA043901 and 2009AA043902.
Rights and permissions
About this article
Cite this article
Wei, H., He, J. & Tan, J. Layered hidden Markov models for real-time daily activity monitoring using body sensor networks. Knowl Inf Syst 29, 479–494 (2011). https://doi.org/10.1007/s10115-011-0423-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10115-011-0423-3