Human activity recognition (HAR) aims to classify and identify activities based on data-driven from different devices, such as sensors or cameras. Particularly, mobile devices have been used for this recognition task. However, versatility of users, location of smartphones, battery, processing and storage limitations, among other issues have been identified. In that sense, this paper presents a human activity recognition system based on artificial hydrocarbon networks. This technique have been proved to be very effective on HAR systems using wearable sensors, so the present work proposes to use this learning method with the information provided by the in-sensors of mobile devices. Preliminary results proved that artificial hydrocarbon networks might be used as an alternative for human activity recognition on mobile devices. In addition, a real dataset created for this work has been published.
CITATION STYLE
Ponce, H., González, G., Miralles-Pechuán, L., & Martínez-Villaseñor, M. L. (2018). Human activity recognition on mobile devices using artificial hydrocarbon networks. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10632 LNAI, pp. 17–29). Springer Verlag. https://doi.org/10.1007/978-3-030-02837-4_2
Mendeley helps you to discover research relevant for your work.