A Study on Human Activity Recognition Using Accelerometer Data from Smartphones

• DOI: 10.1016/j.procs.2014.07.009
• Publication Date: 2014-08-15T14:59:17Z
• Authors: Akram Bayat; Marc Pomplun; Duc A. Tran
• Keywords: Activity Recognition; Data set

This paper describes how to recognize certain types of human physical activities using acceleration data generated by a user's cell phone. We propose recognition system in which new digital low-pass filter is designed order isolate the component gravity from that body raw data. The was trained and tested an experiment with multiple subjects real-world conditions. Several classifiers were various statistical features. High-frequency low-frequency components taken into account. selected five each offering good performance for recognizing our set investigated combine them optimal classifiers. found average probabilities as fusion method could reach overall accuracy rate 91.15%.