The Deep Learning-based Human Activity Recognition Using Smart Wearable Sensors : A Tutorial

Abstract

Human activity recognition (HAR) mechanisms that distinguish human behavior utilizing wearablesensors have advanced significantly over several years. Not only have state-of-the-art techniques ignoredhand-crafted features in favor of end-to-end deep learning approaches, but best practices fordesigning experiments, preparing datasets, and assessing activity recognition systems have changedin lockstep. This tutorial will provide an in-depth, hands-on introduction to the topic of sensor-basedHAR for those who are new to it. We will concentrate on deep learning-based HAR in this tutorialutilizing data from intelligent wearable sensor devices. This tutorial introduces the SDL-HARframework, which provides a general-purpose framework for data preprocessing, data generation,model development, and evaluation. We describe each aspect of the provided system in-depth, offerreferences to relevant research, and explain the community’s best practice methodologies for activityidentification. Two exemplary deep learning approaches, convolutional neural network (CNN) andlong short-term memory neural network (LSTM), are deployed in this lesson using state-of-the-artpublic HAR datasets. Additionally, this tutorial highlights the problems and future research directionsof sensor-based HAR.