Context-Aware Human Activity Recognition (CAHAR) in-the-Wild Using Smartphone Accelerometer

• Published in: IEEE sensors journal Vol. 20; no. 8; pp. 4361 - 4371
• Main Authors: Asim, Yusra, Azam, Muhammad Awais, Ehatisham-ul-Haq, Muhammad, Naeem, Usman, Khalid, Asra
• Format: Journal Article
• Language: English
• Published: New York IEEE 15.04.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: accelerometer; Accelerometers; Activity recognition; Behavior; behavioral context; Biomedical monitoring; context-aware; Environment models; Feature extraction; Human activity recognition; Human performance; Machine learning; Sensor fusion; smartphone; Smartphones; Support vector machines; ubiquitous computing; Wearable sensors
• ISSN: 1530-437X; 1558-1748
• DOI: 10.1109/JSEN.2020.2964278

Smartphones are a promising platform for continuous monitoring of human behavior. However, the ability to capture people's behavioral patterns in-the-wild is a challenge, as the user's behavior and physical activities can vary, given the variability of settings and environments. Modeling and understanding of human activity in-the-wild must not overlook a user's behavioral context, which is just as crucial as recognizing the range of physical activities. The work in this paper presents a novel framework for context-aware human activity recognition by incorporating human behavioral contexts with physical activities. The proposed framework utilizes a series of machine learning classifiers to validate the efficiency of the proposed method.