Online pose classification and walking speed estimation using handheld devices

We describe and evaluate two methods for device pose classification and walking speed estimation that generalize well to new users, compared to previous work. These machine learning based methods are designed for the general case of a person holding a mobile device in an unknown location and require only a single low-cost, low-power sensor: a triaxial accelerometer. We evaluate our methods in straight-path indoor walking experiments as well as in natural indoor walking settings. Experiments with 14 human participants to test user generalization show that our pose classifier correctly selects among four device poses with 94% accuracy compared to 82% for previous work, and our walking speed estimates are within 12-15% (straight/indoor walk) of ground truth compared to 17-22% for previous work. Implementation on a mobile phone demonstrates that both methods can run efficiently online.

[1]  Henk L. Muller,et al.  Personal position measurement using dead reckoning , 2003, Seventh IEEE International Symposium on Wearable Computers, 2003. Proceedings..

[2]  William J. Kaiser,et al.  AutoGait: A mobile platform that accurately estimates the distance walked , 2010, 2010 IEEE International Conference on Pervasive Computing and Communications (PerCom).

[3]  Bahram Honary,et al.  Sensor Placement Modes for Smartphone based Pedestrian Dead Reckoning , 2012 .

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

[5]  Tomaso A. Poggio,et al.  Regularization Networks and Support Vector Machines , 2000, Adv. Comput. Math..

[6]  Eric Foxlin,et al.  Pedestrian tracking with shoe-mounted inertial sensors , 2005, IEEE Computer Graphics and Applications.

[7]  Emiliano Miluzzo,et al.  Pocket, Bag, Hand, etc. - Automatically Detecting Phone Context through Discovery , 2010 .

[8]  FoxlinEric.  Pedestrian Tracking with Shoe-Mounted Inertial Sensors , 2005 .

[9]  Nasser M. Nasrabadi,et al.  Pattern Recognition and Machine Learning , 2006, Technometrics.

[10]  Seth J. Teller,et al.  Growing an organic indoor location system , 2010, MobiSys '10.

[11]  Jun Yang,et al.  Toward physical activity diary: motion recognition using simple acceleration features with mobile phones , 2009, IMCE '09.

[12]  Tae-Seong Kim,et al.  Accelerometer’s position independent physical activity recognition system for long-term activity monitoring in the elderly , 2010, Medical & Biological Engineering & Computing.

[13]  Deborah Estrin,et al.  Using mobile phones to determine transportation modes , 2010, TOSN.

[14]  K. Aminian,et al.  The prediction of speed and incline in outdoor running in humans using accelerometry. , 1999, Medicine and science in sports and exercise.

[15]  Christopher M. Bishop,et al.  Pattern Recognition and Machine Learning (Information Science and Statistics) , 2006 .

[16]  H. Weinberg.  Using the ADXL202 in Pedometer and Personal Navigation Applications , 2002 .

[17]  Michael L. Littman,et al.  Activity Recognition from Accelerometer Data , 2005, AAAI.

[18]  Gunnar Rätsch,et al.  The SHOGUN Machine Learning Toolbox , 2010, J. Mach. Learn. Res..

[19]  Gaurav S. Sukhatme,et al.  Toward free-living walking speed estimation using Gaussian Process-based Regression with on-body accelerometers and gyroscopes , 2010, 2010 4th International Conference on Pervasive Computing Technologies for Healthcare.

[20]  Y. Kawahara,et al.  Recognizing User Context Using Mobile Handsets with Acceleration Sensors , 2007, 2007 IEEE International Conference on Portable Information Devices.

[21]  Alexander J. Smola,et al.  Learning with kernels , 1998 .

[22]  Paul Lukowicz,et al.  Where am I: Recognizing On-body Positions of Wearable Sensors , 2005, LoCA.

[23]  Ian H. Witten,et al.  The WEKA data mining software: an update , 2009, SKDD.

[24]  Gaetano Borriello,et al.  A Practical Approach to Recognizing Physical Activities , 2006, Pervasive.