Activity Recognition Enhancement Based on Ground-Truth: Introducing a New Method Including Accuracy and Granularity Metrics
- 855 Downloads
The uncertainty associated with existing sensing technologies and reasoning methods affects the outcome of the activity recognition process (e.g., accuracy, precision, granularity). The activity recognition process is even challenging when switching from laboratory towards real deployments, where scenarios are not predefined and more complex. Therefore we propose a novel method to improve the activity recognition outcome, by finding a proper balance between accuracy and granularity. The method has been validated through the deployment of UbiSMART (an AAL framework) in 45 scenarios of ageing in place. We discuss in this paper our method and the validation results.
KeywordsAmbient assisted living Activity recognition Semantic reasoning Quality insurance Ground-truth acquisition
This research project has been supported by the Quality Of Life Chair supported by Foundation Telecom of the Institut Mines-Telecom in France, La Mutuelle Generale and REUNICA which figure among the major health-care insurance companies in France. The work is also supported by the grand emprunt VHP inter@ctive project. We also wish to acknowledge the support of the Saint-Vincent-de-Paul nursing home and its director Brigitte Choquet, who kindly let us deploy our system within their environment.
- 2.Basili, V.R.: Software modeling and measurement: the goal/question/metric paradigm (1992)Google Scholar
- 3.Biswas, J., et al.: Activity recognition in assisted living facilities with incremental, approximate ground truth. In: Geissbühler, A., Demongeot, J., Mokhtari, M., Abdulrazak, B., Aloulou, H. (eds.) ICOST 2015. LNCS, vol. 9102, pp. 103–115. Springer, Cham (2015). doi: 10.1007/978-3-319-19312-0_9 CrossRefGoogle Scholar
- 5.Cook, D.J.: Learning setting-generalized activity models for smart spaces. IEEE Intell. Syst. 2010(99), 1 (2010)Google Scholar
- 6.Fillenbaum, G.G.: Multidimensional Functional Assessment of Older Adults: The Duke Older Americans Resources and Services procedures. Psychology Press, London (2013)Google Scholar
- 7.Jahn, A., David, K.: Improved activity recognition by using grouped activities. In: 2016 IEEE International Conference on Pervasive Computing and Communication Workshops (PerCom Workshops), pp. 1–5. IEEE (2016)Google Scholar
- 8.Jokela, T., Iivari, N., Matero, J., Karukka, M.: The standard of user-centered design and the standard definition of usability: analyzing ISO 13407 against ISO 9241–11. In: Proceedings of the Latin American Conference on Human-Computer Interaction, pp. 53–60. ACM (2003)Google Scholar
- 9.Kadouche, R., Abdulrazak, B., Mokhtari, M., Giroux, S., Pigot, H.: Semantic matching framework for handicap situation detection in smart environments. J. Ambient Intell. Smart Environ. 1(3), 223–234 (2009)Google Scholar
- 11.Kleinberger, T., Jedlitschka, A., Storf, H., Steinbach-Nordmann, S., Prueckner, S.: An approach to and evaluations of assisted living systems using ambient intelligence for emergency monitoring and prevention. In: Stephanidis, C. (ed.) UAHCI 2009. LNCS, vol. 5615, pp. 199–208. Springer, Heidelberg (2009). doi: 10.1007/978-3-642-02710-9_23 CrossRefGoogle Scholar
- 13.Lawton, M.P.: Scales to measure competence in everyday activities. Psychopharmacol. Bull. 24(4), 609–614 (1987)Google Scholar
- 16.Slim, S.O., Atia, A., Mostafa, M.-S.M.: An experimental comparison between seven classification algorithms for activity recognition. In: Gaber, T., Hassanien, A.E., El-Bendary, N., Dey, N. (eds.) AISI 2015. AISC, vol. 407, pp. 37–46. Springer, Cham (2016). doi: 10.1007/978-3-319-26690-9_4 Google Scholar
- 17.Yared, R., Abdulrazak, B., Tessier, T., Mabilleau, P.: Cooking risk analysis to enhance safety of elderly people in smart kitchen. In: Proceedings of the 8th ACM International Conference on Pervasive Technologies Related to Assistive Environments, p. 12. ACM (2015)Google Scholar