Empirical Study of Design Choices in Multi-Sensor Context Recognition Systems

This paper deals with the design and implementation of a highlyminiaturized, multi-sensor context recognition sy stem. It repre- sents an optimal trade-off between power consumption and recognition performance rather than straightforward maximization of the recognition rate. We present a thumb-sized, 8 gram platform that combines sound, accel- eration and light sensing with processing power, wireless communication and a battery. Based on this platform we make an experimental evalu- ation of design choices present in such multi-sensor context recognition systems. We introduce a design method to achieve an optimal power con- sumption vs. recognition rate trade-off through variations of the sampling rate, feature selection and choice of classifiers. Power consumption anal- ysis indicates that our system can operate for 300 hours without having to recharge the battery. An important and somewhat surprising result of our analysis is that the addition of a sensor maybe a power efficient wayto improve the overall system performance.

[1]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[2]  Daniel P. Siewiorek,et al.  System design approach to power aware mobile computers , 2003, IEEE Computer Society Annual Symposium on VLSI, 2003. Proceedings..

[3]  Paul Lukowicz,et al.  Implementation and evaluation of a low-power sound-based user activity recognition system , 2004, Eighth International Symposium on Wearable Computers.

[4]  Ian H. Witten,et al.  Data mining: practical machine learning tools and techniques with Java implementations , 2002, SGMD.

[5]  Paul Lukowicz,et al.  A systematic approach to the design of distributed wearable systems , 2004, IEEE Transactions on Computers.

[6]  Paul Lukowicz,et al.  Sampling frequency, signal resolution and the accuracy of wearable context recognition systems , 2004, Eighth International Symposium on Wearable Computers.

[7]  Bernt Schiele,et al.  A model for human interruptability: experimental evaluation and automatic estimation from wearable sensors , 2004, Eighth International Symposium on Wearable Computers.

[8]  E. Lindsay,et al.  Chapter 10 , 1994, Genji & Heike.

[9]  Ling Bao,et al.  Activity Recognition from User-Annotated Acceleration Data , 2004, Pervasive.

[10]  Gregory D. Abowd,et al.  Context-awareness in wearable and ubiquitous computing , 1997, Digest of Papers. First International Symposium on Wearable Computers.

[11]  C. Randell,et al.  Context awareness by analysing accelerometer data , 2000, Digest of Papers. Fourth International Symposium on Wearable Computers.

[12]  T. Tamura,et al.  Classification of walking pattern using acceleration waveform in elderly people , 2000, Proceedings of the 22nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (Cat. No.00CH37143).

[13]  Paul Lukowicz,et al.  Recognizing Workshop Activity Using Body Worn Microphones and Accelerometers , 2004, Pervasive.

[14]  Paul Lukowicz,et al.  PadNET: wearable physical activity detection network , 2003, Seventh IEEE International Symposium on Wearable Computers, 2003. Proceedings..

[15]  Albrecht Schmidt,et al.  Building intelligent environments with Smart-Its , 2004, IEEE Computer Graphics and Applications.

[16]  Gerhard Tröster,et al.  Towards Wearable Autonomous Microsystems , 2004, Pervasive.

[17]  Paul Lukowicz,et al.  WearNET: A Distributed Multi-sensor System for Context Aware Wearables , 2002, UbiComp.

[18]  M. Hannikainen,et al.  Wireless sensor prototype platform , 2003, IECON'03. 29th Annual Conference of the IEEE Industrial Electronics Society (IEEE Cat. No.03CH37468).

[19]  Ishwar K. Sethi,et al.  Classification of general audio data for content-based retrieval , 2001, Pattern Recognit. Lett..

[20]  Tom Fawcett,et al.  Analysis and Visualization of Classifier Performance: Comparison under Imprecise Class and Cost Distributions , 1997, KDD.

[21]  Bernt Schiele,et al.  Towards a wearable inertial sensor network , 2003 .

[22]  Alberto Maria Segre,et al.  Programs for Machine Learning , 1994 .

[23]  Paul Lukowicz,et al.  SoundButton: design of a low power wearable audio classification system , 2003, Seventh IEEE International Symposium on Wearable Computers, 2003. Proceedings..

[24]  Kristof Van Laerhoven,et al.  Context awareness in Systems with Limited Resources , 2002 .