Transformer With Bidirectional GRU for Nonintrusive, Sensor-Based Activity Recognition in a Multiresident Environment

Several techniques for human activity recognition (HAR) in a smart indoor environment have been developed and improved along with the rapid advancement of sensor technologies. However, recognizing multiple people’s activities is still challenging due to the complexity of their activities, such as parallel and collaborative activities. To address these challenges, we propose a transformer with a bidirectional gated recurrent unit (GRU) deep learning (DL) method, called TRANS-BiGRU, to efficiently learn and recognize different types of activities performed by multiple residents. We compare the proposed model with the state-of-the-art models and various DL models, such as Ensemble2LSTM (Ens2-LSTM), bidirectional GRUs (Bi-GRU), and traditional machine learning (ML) models, such as support vector machine (SVM). Our experimental results based on the center for advanced studies in adaptive system and ARAS public data sets show that our model significantly outperforms the existing models for complex activity recognition of multiple residents.

I. Introduction

Our global population is entering an aging stage, and the number and proportion of older people in almost every country in the world are fast increasing. By 2050, the number of people aged 60 will reach 2.1 billion [1]. As they grow older, many of them are likely to suffer from some geriatric diseases, such as dementia or Alzheimer’s, which will make them more challenging to live independently [2]. Due to the shortage of caregivers and the imperfection of available healthcare infrastructure, many studies have begun to establish the assisted living system to help monitor activities of daily living (ADLs) [3], [4]. Especially, human activity recognition (HAR) is a vital research area that focuses on identifying human actions based on the information obtained from various kinds of sensors used for activity monitoring [5].