Background - Smartphone-Based Human Activity Recognition - page 30

•

Battery consumption : quantifies the energy expenditure of portable devices and
how this affects their continuous operation time.

• Memory and CPU usage : considers limitations regarding memory and CPU

requirements. It is critical for instance in systems that share resources with others
applications (e.g. in smartphones, personal computers).

• Obstrusivity : evaluates how comfortable is the system for the user (e.g. sensors

location and weight, presence of wires, fitting time, etc.)

• User's privacy : examines whether the system protects the accessed personal data

of its users from external sources.

• Number and type of classes the system is able to classify.

• Number and type of sensors required for the recognition of activities.

• Modular design : indicates whether or not the system allows its integration with

others or the adaptation of new sensors and devices.

2.6 Summary

In this chapter we described the background required to contextualize the problem
of HAR. We first covered the fields of application our research work is targeting to:
AAL and AmI which are human-centered areas aiming to improve people's QoL
through the use of smart technologies. Then, we explored three implementation
considerations that constitute themain building blocks involved in the development of
our proposedHAR systems. These are: Sensing devices, with focus on inertial sensors
(accelerometers and gyroscope), for the detection of physical activity; smartphones,
the recent everyday use device with strong computing capabilities now exploited as
a novel service provider (HAR in our case); and the intelligence behind our systems
provided by SVMs.

References

E. Aarts, R. Wichert, in Ambient Intelligence . Technology Guide, 2009
G.D. Abowd, E.D. Mynatt, Charting past, present, and future research in ubiquitous computing.
ACM Trans. Comput. Hum. Interact. 7 , 29-58 (2000)
A. Aizerman, E.M. Braverman, L.I. Rozoner, Theoretical foundations of the potential function
method in pattern recognition learning. Autom. Remote Control 25 , 821-837 (1964)
N.S. Altman, An introduction to kernel and nearest-neighbor nonparametric regression. Am. Stat.
46 , 175-185 (1992)
D. Anguita, A. Ghio, L. Oneto, X. Parra, J.-L. Reyes-Ortiz, in Training Computationally Efficient
Smartphone-Based Human Activity Recognition Models . International Conference on Artificial
Neural Networks, 2013
X. Bian, G.D Abowd, J.M Rehg, Using sound source localization in a home environment. in Per-
vasive Computing (2005)
C.M. Bishop. Pattern Recognition and Machine Learning (Springer, Heidelberg, 2006)