SIT4Energy is delighted to announce its first international journla publication. Research results on human behaviour, entitled “Real-time Physical Activity Recognition on Smart Mobile Devices using Convolutional Neural Networks”, authored by Kostas Peppas, Apostolos C. Tsolakis, Stelios Krinidis, and Dimitrios Tzovaras, has been accepted for publication today!.
It presents the results of the novel AI-driven architecture developed by CERTH for activity tracking on a mobile device. Results demonstrate improved performance when compared with other state-of-the-art solutions on two different datasets.
In the context of SIT4Energy this tool enriches the micro-moments engine for identifying the optimal moment to sent a recommendation to the end-user.
Find the open access publication here.