DNN-Based Approach for Recognition of Human Activity Raw Data in Non-Controlled Environment
Hamdi Amroun, M’hamed Temkit, Mehdi Ammi · 2017
In this paper, we ask whether accurate recognition of activity can be obtained by using a network of smart objects. The approach consists in the classification of certain activities of the subjects: walking, standing, sitting and lying down. The study uses a network of commonly connected objects: a smart watch, a smartphone and a remote control and transported by the participants during an uncontrolled experiment. The sensor data of the three devices were classified by a deep neural networks (DNN) algorithm without prior pre-processing of the data. We show that (DNN) provides better results compared to Decision Tree (DT) and Support Vector Machine (SVM) algorithms. The results also show that the activities of the participants were classified with an accuracy of more than 98.53%, on average.