Deep Neural Network Based Model for Context-Aware Human Activity Recognition

Abstract

currently, with the growth of smart sensing technologies in ubiquitous computing. Human activity recognition (HAR) is becoming a fundamental research problem. HAR aims to recognize the person’s body position, motion, and function with camera and sensors-based systems. Even though camera based HAR gain much progress but due to certain privacy concerns researchers focus with cost-effective sensor-based miniatures for HAR. Because it can play a vital role in aging care, smart homes, and daily life assistant Apps. As the human activities bring a lot of information about context that can help models to accomplish context-awareness. The precise acknowledgement of in¬the¬wild human activities and the contexts related with these activities remains an open research challenge that needs to be addressed. In this work, the aim is to present a context aware human activity recognition (CAHAR) scheme to learn the variability of human behavior context in the wild with physical activity recognition. Deep neural networks and Machine Learning (ML) algorithms opted to get behavioral context of a person in the designed scheme of CAHAR and use different machine learning classifier for comparison with the presented scheme