Browse DORAS
Browse Theses
Search
Latest Additions
Creative Commons License
Except where otherwise noted, content on this site is licensed for use under a:

Can future physical assessment continue without support from computer science?

Doherty, Aiden R. and Kelly, Paul and Smeaton, Alan F. and Foster, Charlie (2011) Can future physical assessment continue without support from computer science? In: ISBNPA2011: the 2011 Annual Meeting of the International Society for Behavioral Nutrition and Physical Activity, 5-18 June 2011, Melbourne, Australia.

Full text available as:

[img]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
3040Kb

Abstract

Purpose: Public health research requires data-intense studies over extended periods. With the advent of new technologies, there has been a resulting explosion in the amount of data generated by wearable sensors that can be used in physical activity research. Unfortunately the advances in hardware (e.g. device size), have not been matched by software to help manage, organise and analyse this data deluge. Public health research will require cross-disciplinary interactions with the computer science community in working towards solutions to automatically recognise human activities from wearable sensor data. Methods: We conducted a meta-review of contemporary computing science and information retrieval approaches such as: 1) the management and indexing of data from wearable accelerometer and image capturing devices; 2) the synchronisation and fusion of data from multiple devices (e.g. GPS, accelerometer, & SenseCam data); and 3) the representation of the meaning of image data Results: We identified: 1) Relational databases offer the most flexible solution for managing, indexing, and querying wearable sensor data information. 2) Advanced analysis of the signature of streams of data from separate devices may allow their data to be synchronised, but requires further advances. 3) The optimum method of pattern recognition in automatically annotating wearable image data is through the use of Support Vector Machine classifiers. Conclusions: Close interdisciplinary research between public health and computing science can help further our understanding of human activities. A first step may be the use of sensor web technologies to recognise a well-defined subset of activities that individuals are engaged in.

Item Type:Conference or Workshop Item (Speech)
Event Type:Conference
Refereed:No
Additional Information:This research is funded by the Irish Health Research Board under grant number MCPD/2010/12. This material is also based upon works supported by the Science Foundation Ireland under Grant No. 07/CE/I1147 and the British Heart Foundation. In addition this work was supported by Microsoft Research through its PhD Scholarship Programme.
Uncontrolled Keywords:Data management; SenseCam activity recognition
Subjects:Computer Science > Lifelog
Medical Sciences > Exercise
Computer Science > Computer software
Medical Sciences > Epidemiology
Computer Science > Multimedia systems
Computer Science > Information retrieval
Medical Sciences > Sports sciences
DCU Faculties and Centres:Research Initiatives and Centres > Centre for Digital Video Processing (CDVP)
Research Initiatives and Centres > CLARITY: The Centre for Sensor Web Technologies
DCU Faculties and Schools > Faculty of Engineering and Computing > School of Computing
Use License:This item is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 3.0 License. View License
Funders:Science Foundation Ireland, Health Research Board, British Heart Foundation, Microsoft Research
ID Code:16836
Deposited On:09 Feb 2012 11:49 by Aiden Doherty. Last Modified 09 Feb 2012 11:49

Download statistics

Archive Staff Only: edit this record