Abstract
A common assumption in activity recognition is that the system remains unchanged between its design and its posterior operation. However, many factors affect the data distribution between two different experimental sessions. One of these factors is the potential change in the sensor location (e.g. due to replacement or slippage) affecting the classification performance. Assuming that changes in the sensor placement mainly result in shifts in the feature distributions, we propose an unsupervised adaptive classifier that calibrates itself using an online version of expectation–maximisation. Tests using three activity recognition scenarios show that the proposed adaptive algorithm is robust against shift in the feature space due to sensor displacement and rotation. Moreover, since the method estimates the change in the feature distribution, it can also be used to roughly evaluate the reliability of the system during online operation.
Similar content being viewed by others
References
Alippi C, Roveri M (2006) An adaptive cusum-based test for signal change detection. In: ISCAS. IEEE
Bayati H, Millán JR, Chavarriaga R (2011) Unsupervised adaptation to on-body sensor displacement in acceleration-based activity recognition. In: IEEE international symposium on wearable computers, ISWC
Bishop CM (2007) Pattern recognition and machine learning. Springer, New York
Chandola V, Banerjee A, Kumar V (2009) Anomaly detection: a survey. ACM Comput Surv 41(3):1–58
Chavarriaga R, Sagha H, Millán JR (2011) Ensemble creation and reconfiguration for activity recognition: an information theoretic approach. In: IEEE international conference on systems, man, and cybernetics (IEEE SMC)
Duda RO, Hart PE, Stork DG (2000) Pattern classification. 2nd edn. Wiley-Interscience, New York
Förster K, Brem P, Roggen D, Tröster G (2009) Evolving discriminative features robust to sensor displacement for activity recognition in body area sensor networks. In: International conference intelligent sensors, sensor networks and information processing, ISSNIP pp 43–48
Förster K, Roggen D, Tröster G (2009) Unsupervised classifier self-calibration through repeated context occurences: is there robustness against sensor displacement to gain? In: IEEE international symposium on wearable computers
Hasanuzzaman M, Zhang T, Ampornaramveth V, Gotoda H, Shirai Y, Ueno H (2007) Adaptive visual gesture recognition for human-robot interaction using a knowledge-based software platform. Robot Auton Syst 55(8):643–657
Kang H, Lee CW, Jung K (2004) Recognition-based gesture spotting in video games. Pattern Recogn Lett 25(15):1701–1714
Krzanowski WJ (1998) Principles of multivariate analysis. Oxford University Press, Oxford
Kunze K, Lukowicz P (2008) Dealing with sensor displacement in motion-based onbody activity recognition systems. In: Proceedings of the international conference on Ubiquitous computing, pp 20–29. ACM, New York, NY, USA
Kurz M, Hölzl G, Ferscha A, Calatroni A, Roggen D, Tröster G, Sagha H, Chavarriaga R, Millán JR, Bannach D, Kunze K, Lukowicz P (2011) The opportunity framework and data processing ecosystem for opportunistic activity and context recognition. Int J Sens Wirel Commun Control (in press)
Lester J, Choudhury T, Borriello G (2006) A practical approach to recognizing physical activities. In: Lecture notes in computer science: pervasive computing
Li Y, Kambara H, Koike Y, Sugiyama M (2010) Application of covariate shift adaptation techniques in brain computer interface. IEEE Trans Biomed Eng 57(6):1318–1324
Lukowicz P, Pirkl G, Bannach D, Wagner F, Calatroni A, Förster K, Holleczek T, Rossi M, Roggen D, Tröster G, Doppler J, Holzmann C, Riener A, Ferscha A, Chavarriaga R (2010) Recording a complex, multi modal activity data set for context recogntion. In: 1st Workshop on context-systems design, evaluation and optimisation at ARCS
Manly BFJ, Mackenzie D (2000) A cumulative sum type of method for environmental monitoring. Environmetrics 11(2):151–166
Marquardt DW (1963) An algorithm for least-squares estimation of nonlinear parameters. J Soc Ind Appl Math 11(2):431–441
Ravi N, D N, Mysore P, Littman ML (2005) Activity recognition from accelerometer data. In: Proceedings of the 17th conference on innovative applications of artificial intelligence 3
Roggen D, Calatroni A, Rossi M, Holleczek T, Förster K, Tröster G, Lukowicz P, Bannach D, Pirkl G, Ferscha A, Doppler J, Holzmann C, Kurz M, Holl G, Chavarriaga R, Sagha H, Bayati H, Creatura M, Millán JR (2010) Collecting complex activity data sets in highly rich networked sensor environments. In: Seventh international conference on networked sensing systems
Roggen D, Förster K, Calatroni A, Holleczek T, Fang Y, Tröster G, Lukowicz P, Pirkl G, Bannach D, Kunze K, Ferscha A, Holzmann C, Riener A, Chavarriaga R, Millán J (2009) OPPORTUNITY: towards opportunistic activity and context recognition systems. In: Third IEEE WoWMoM workshop on autonomic and opportunistic communications
Sagha H, Millán, JR, Chavarriaga R (2011) Detecting and rectifying anomalies in opportunistic sensor networks. In: International conference on body sensor networks, BSN. Dallas, Texas, USA
Sagha H, Millán JR, Chavarriaga R (2011) Detecting anomalies to improve classification performance in an opportunistic sensor network. In: IEEE workshop on sensor networks and systems for pervasive computing, PerSens. Seattle
Smyth P (1994) Hidden Markov models for fault detection in dynamic systems. Pattern Recogn 27(1):149–164
Steinhoff U, Schiele B (2010) Dead reckoning from the pocket—an experimental study. In: IEEE international conference on pervasive computing and communications (PerCom 2010). Mannheim, Germany
Stiefmeier T, Roggen D, Tröster G, Ogris G, Lukowicz P (2008) Wearable activity tracking in car manufacturing. IEEE Pervas Comput 7(2):42–50
Sugiyama M, Krauledat M, Müller KR (2007) Covariate shift adaptation by importance weighted cross validation. J Mach Learn Res 8:985–1005
Sugiyama M, Suzuki T, Nakajima S, Kashima H, von Bnau P, Kawanabe M (2008) Direct importance estimation for covariate shift adaptation. Ann Inst Stat Math 60(4):699–746
Tentori M, Favela J (2008) Activity-aware computing for healthcare. IEEE Pervas Comput 7(2):51–57
Van Laerhoven K, Cakmakci O (2000) What shall we teach our pants? In: IEEE international symposium on wearable computers, IWSC pp 77–83
Westeyn T, Vadas K, Bian X, Starner T, Abowd G (2005) Recognizing mimicked autistic self-stimulatory behaviors using HMMs. In: Ninth IEEE international symposium on wearable computers proceedings, pp 164–167
Wilson AD, Bobick AF (2000) Realtime online adaptive gesture recognition. In: Pattern recognition, international conference on 1, 1270
Acknowledgments
We would like to thank K. Förster and D. Roggen from ETH Zurich, and H. Sagha from EPFL, Lausanne for providing the experimental data and insightful discussions. This work was supported by the EU-FET project ICT-225938 (Opportunity: Activity and Context Recognition with Opportunistic Sensor Configuration). This paper only reflects the authors’ views and funding agencies are not liable for any use that may be made of the information contained herein.
Author information
Authors and Affiliations
Corresponding author
Additional information
Based on “Unsupervised adaptation to on-body sensor displacement in acceleration-based activity recognition”, by H. Bayati, J. R. Millán, and R. Chavarriaga which appeared in the Proceedings of the International Symposium on Wearable Computers, San Francisco, California, June 2011. © 2011 IEEE.
R. Chavarriaga and H. Bayati contributed equally to this work.
Rights and permissions
About this article
Cite this article
Chavarriaga, R., Bayati, H. & Millán, J.R. Unsupervised adaptation for acceleration-based activity recognition: robustness to sensor displacement and rotation. Pers Ubiquit Comput 17, 479–490 (2013). https://doi.org/10.1007/s00779-011-0493-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00779-011-0493-y